diff --git a/app/(admin)/settings/admin-settings-client.tsx b/app/(admin)/settings/admin-settings-client.tsx
index 5a44689c4..4f9a3ac5b 100644
--- a/app/(admin)/settings/admin-settings-client.tsx
+++ b/app/(admin)/settings/admin-settings-client.tsx
@@ -396,6 +396,13 @@ export function AdminSettingsClient() {
     timeout: 120,
   });
 
+  // 场景裂变配置状态
+  const [sceneFusionConfig, setSceneFusionConfig] = useState({
+    enabled: true,
+    defaultPrompt: "将图1的产品替换到图2的场景中，保持图1产品的外观、颜色和细节不变，自然融合到图2的环境中。",
+    maxScenes: 10,
+  });
+
   // 加载抠图API配置
   const { data: rmbgConfigData } = useSWR("/api/admin/settings?key=rmbg_api_config", fetcher);
   useEffect(() => {
@@ -479,6 +486,23 @@ export function AdminSettingsClient() {
     }
   }, [angleEditConfigData]);
 
+  // 加载场景裂变配置
+  const { data: sceneFusionConfigData } = useSWR("/api/admin/settings?key=scene_fusion_config", fetcher);
+  useEffect(() => {
+    if (sceneFusionConfigData?.setting?.value) {
+      const config = sceneFusionConfigData.setting.value as { 
+        enabled?: boolean;
+        defaultPrompt?: string;
+        maxScenes?: number;
+      };
+      setSceneFusionConfig({
+        enabled: config.enabled ?? true,
+        defaultPrompt: config.defaultPrompt || "将图1的产品替换到图2的场景中，保持图1产品的外观、颜色和细节不变，自然融合到图2的环境中。",
+        maxScenes: config.maxScenes ?? 10,
+      });
+    }
+  }, [sceneFusionConfigData]);
+
   // 导航链接状态
   const [navLinkForm, setNavLinkForm] = useState({
     id: "",
@@ -922,6 +946,34 @@ export function AdminSettingsClient() {
     }
   };
 
+  // --- 场景裂变配置逻辑 ---
+  const handleSaveSceneFusionConfig = async () => {
+    try {
+      if (!sceneFusionConfig.defaultPrompt.trim()) {
+        toast.error("默认提示词不能为空");
+        return;
+      }
+
+      const response = await fetch("/api/admin/settings", {
+        method: "POST",
+        headers: { "Content-Type": "application/json" },
+        body: JSON.stringify({
+          key: "scene_fusion_config",
+          value: {
+            enabled: sceneFusionConfig.enabled,
+            defaultPrompt: sceneFusionConfig.defaultPrompt.trim(),
+            maxScenes: sceneFusionConfig.maxScenes,
+          },
+        }),
+      });
+
+      if (!response.ok) throw new Error("保存失败");
+      toast.success("场景裂变配置保存成功");
+    } catch (error) {
+      toast.error("保存场景裂变配置失败");
+    }
+  };
+
   // --- 导航链接逻辑 ---
   const resetNavLinkForm = () => {
     setNavLinkForm({ id: "", name: "", url: "", order: 0, isEnabled: true });
@@ -3764,6 +3816,81 @@ export function AdminSettingsClient() {
               </Button>
             </CardContent>
           </Card>
+
+          {/* 场景裂变配置 */}
+          <Card>
+            <CardHeader>
+              <CardTitle className="flex items-center gap-2">
+                <Sparkles className="h-5 w-5" />
+                场景裂变配置
+              </CardTitle>
+              <CardDescription>
+                配置场景裂变功能的默认提示词和参数。场景裂变可将产品图批量融合到多个场景中。
+              </CardDescription>
+            </CardHeader>
+            <CardContent className="space-y-6 pt-6">
+              {/* 功能开关 */}
+              <div className="flex items-center justify-between rounded-lg border p-4">
+                <div className="space-y-0.5">
+                  <Label className="text-base">启用场景裂变功能</Label>
+                  <div className="text-sm text-muted-foreground">
+                    开启后用户可在图片编辑器中使用场景裂变功能
+                  </div>
+                </div>
+                <Switch
+                  checked={sceneFusionConfig.enabled}
+                  onCheckedChange={(checked) => setSceneFusionConfig({ ...sceneFusionConfig, enabled: checked })}
+                />
+              </div>
+
+              {/* 默认提示词 */}
+              <div className="space-y-2">
+                <Label htmlFor="sceneFusionPrompt">默认提示词 *</Label>
+                <Textarea
+                  id="sceneFusionPrompt"
+                  value={sceneFusionConfig.defaultPrompt}
+                  onChange={(e) => setSceneFusionConfig({ ...sceneFusionConfig, defaultPrompt: e.target.value })}
+                  placeholder="输入默认的融合提示词..."
+                  rows={4}
+                  disabled={!sceneFusionConfig.enabled}
+                />
+                <div className="text-xs text-muted-foreground">
+                  该提示词将作为场景裂变的基础指令，用户可以在此基础上补充关键词
+                </div>
+              </div>
+
+              {/* 最大场景数量 */}
+              <div className="space-y-2">
+                <Label htmlFor="maxScenes">最大场景数量</Label>
+                <Input
+                  id="maxScenes"
+                  type="number"
+                  value={sceneFusionConfig.maxScenes}
+                  onChange={(e) => setSceneFusionConfig({ ...sceneFusionConfig, maxScenes: Number.parseInt(e.target.value) || 10 })}
+                  min={1}
+                  max={20}
+                  disabled={!sceneFusionConfig.enabled}
+                />
+                <div className="text-xs text-muted-foreground">
+                  用户单次裂变可上传的最大场景图数量（1-20张）
+                </div>
+              </div>
+
+              <div className="rounded-lg border border-blue-200 bg-blue-50 p-4 space-y-2">
+                <div className="text-sm font-medium text-blue-900">功能说明</div>
+                <div className="text-xs text-blue-800 space-y-1">
+                  <p>• 用户上传1张产品图和多张场景参考图</p>
+                  <p>• 系统自动将产品图融合到每个场景中</p>
+                  <p>• 使用已配置的阿里生图模型进行处理</p>
+                  <p>• 生成的图片会自动保存到用户相册</p>
+                </div>
+              </div>
+              
+              <Button onClick={handleSaveSceneFusionConfig}>
+                <Save className="mr-2 h-4 w-4" /> 保存配置
+              </Button>
+            </CardContent>
+          </Card>
         </TabsContent>
 
       </Tabs>
diff --git a/app/api/scene-fusion/config/route.ts b/app/api/scene-fusion/config/route.ts
new file mode 100644
index 000000000..37b14ba87
--- /dev/null
+++ b/app/api/scene-fusion/config/route.ts
@@ -0,0 +1,55 @@
+/**
+ * 场景裂变配置 API（公开端点）
+ * 
+ * 允许普通用户获取场景裂变的配置信息
+ */
+
+import { NextResponse } from "next/server";
+import { auth } from "@/app/(auth)/auth";
+import { getSystemSetting } from "@/lib/db/queries";
+
+// 默认提示词
+const FALLBACK_PROMPT = "将图1的产品替换到图2的场景中，保持图1产品的外观、颜色和细节不变，自然融合到图2的环境中。";
+
+export async function GET() {
+  try {
+    // 需要登录但不需要 admin 权限
+    const session = await auth();
+    if (!session?.user?.id) {
+      return NextResponse.json(
+        { error: "未登录" },
+        { status: 401 }
+      );
+    }
+
+    // 获取配置
+    const setting = await getSystemSetting({ key: "scene_fusion_config" });
+    
+    if (setting?.value) {
+      const config = setting.value as {
+        enabled?: boolean;
+        defaultPrompt?: string;
+        maxScenes?: number;
+      };
+      
+      return NextResponse.json({
+        enabled: config.enabled ?? true,
+        defaultPrompt: config.defaultPrompt || FALLBACK_PROMPT,
+        maxScenes: config.maxScenes ?? 10,
+      });
+    }
+
+    // 返回默认配置
+    return NextResponse.json({
+      enabled: true,
+      defaultPrompt: FALLBACK_PROMPT,
+      maxScenes: 10,
+    });
+  } catch (error) {
+    console.error("[SceneFusion Config] Error:", error);
+    return NextResponse.json(
+      { error: "获取配置失败" },
+      { status: 500 }
+    );
+  }
+}
diff --git a/app/api/scene-fusion/route.ts b/app/api/scene-fusion/route.ts
new file mode 100644
index 000000000..00d144f45
--- /dev/null
+++ b/app/api/scene-fusion/route.ts
@@ -0,0 +1,406 @@
+/**
+ * 场景裂变 API
+ * 
+ * 将一张产品图批量融合到多个场景图中
+ * 使用阿里云千问图片编辑模型
+ */
+
+import { NextResponse } from "next/server";
+import { z } from "zod";
+import { auth } from "@/app/(auth)/auth";
+import { db } from "@/lib/db";
+import { generatedImage } from "@/lib/db/schema";
+import { getEnabledAiModelsWithProvider, getSystemSetting } from "@/lib/db/queries";
+import { uploadFile } from "@/lib/storage/service";
+import sharp from "sharp";
+
+// 请求体 Schema
+const sceneFusionRequestSchema = z.object({
+  productImage: z.string().min(1, "产品图不能为空"),
+  scenes: z.array(z.object({
+    id: z.string(),
+    image: z.string().min(1, "场景图不能为空"),
+    customPrompt: z.string().optional(),
+  })).min(1, "至少需要一个场景图").max(20, "最多支持20个场景图"),
+});
+
+// 默认提示词（当数据库未配置时使用）
+const FALLBACK_PROMPT = "将图1的产品替换到图2的场景中，保持图1产品的外观、颜色和细节不变，自然融合到图2的环境中。";
+
+/**
+ * 获取场景裂变配置
+ */
+async function getSceneFusionConfig() {
+  try {
+    const setting = await getSystemSetting({ key: "scene_fusion_config" });
+    if (setting?.value) {
+      const config = setting.value as {
+        enabled?: boolean;
+        defaultPrompt?: string;
+        maxScenes?: number;
+      };
+      return {
+        enabled: config.enabled ?? true,
+        defaultPrompt: config.defaultPrompt || FALLBACK_PROMPT,
+        maxScenes: config.maxScenes ?? 10,
+      };
+    }
+  } catch (error) {
+    console.error("[SceneFusion] Failed to load config:", error);
+  }
+  
+  return {
+    enabled: true,
+    defaultPrompt: FALLBACK_PROMPT,
+    maxScenes: 10,
+  };
+}
+
+/**
+ * 创建缩略图
+ */
+async function createThumbnail(buffer: Buffer): Promise<Buffer> {
+  try {
+    const thumbnailBuffer = await sharp(buffer)
+      .resize(400, 400, { fit: "inside", withoutEnlargement: true })
+      .jpeg({ quality: 80 })
+      .toBuffer();
+    return thumbnailBuffer;
+  } catch (error) {
+    console.error("[SceneFusion] Failed to create thumbnail:", error);
+    return buffer;
+  }
+}
+
+/**
+ * 生成唯一文件名
+ */
+function generateFilename(prefix: string, extension: string): string {
+  const timestamp = Date.now();
+  const random = Math.random().toString(36).substring(2, 8);
+  return `${prefix}-${timestamp}-${random}.${extension}`;
+}
+
+/**
+ * 将 Base64 图片上传到 S3 获取 URL
+ */
+async function uploadBase64ToS3(base64Data: string, prefix: string = "scene-fusion"): Promise<string> {
+  try {
+    // 移除 data URL 前缀
+    const base64Clean = base64Data.replace(/^data:image\/\w+;base64,/, "");
+    const buffer = Buffer.from(base64Clean, "base64");
+    
+    const filename = generateFilename(`${prefix}-ref`, "png");
+    const result = await uploadFile(filename, buffer.buffer.slice(buffer.byteOffset, buffer.byteOffset + buffer.byteLength));
+    console.log(`[SceneFusion] Uploaded reference image to: ${result.url}`);
+    return result.url;
+  } catch (error) {
+    console.error("[SceneFusion] Failed to upload reference image:", error);
+    throw error;
+  }
+}
+
+/**
+ * 下载图片并上传到 S3
+ */
+async function downloadAndUploadToS3(
+  imageUrl: string,
+  prefix: string = "scene-fusion"
+): Promise<{ thumbnail: string; original: string }> {
+  try {
+    console.log("[SceneFusion] Downloading image from:", imageUrl.substring(0, 100) + "...");
+    
+    const response = await fetch(imageUrl);
+    if (!response.ok) {
+      throw new Error(`Failed to download image: ${response.status}`);
+    }
+    
+    const buffer = Buffer.from(await response.arrayBuffer());
+    console.log(`[SceneFusion] Downloaded image, size: ${buffer.length} bytes`);
+    
+    // 创建缩略图
+    const thumbnailBuffer = await createThumbnail(buffer);
+    
+    // 上传原图
+    const originalFilename = generateFilename(`${prefix}-orig`, "png");
+    const originalResult = await uploadFile(originalFilename, buffer.buffer.slice(buffer.byteOffset, buffer.byteOffset + buffer.byteLength));
+    console.log(`[SceneFusion] Uploaded original to: ${originalResult.url}`);
+    
+    // 上传缩略图
+    const thumbnailFilename = generateFilename(`${prefix}-thumb`, "jpg");
+    const thumbnailArrayBuffer = new Uint8Array(thumbnailBuffer).buffer;
+    const thumbnailResult = await uploadFile(thumbnailFilename, thumbnailArrayBuffer);
+    console.log(`[SceneFusion] Uploaded thumbnail to: ${thumbnailResult.url}`);
+    
+    return {
+      original: originalResult.url,
+      thumbnail: thumbnailResult.url,
+    };
+  } catch (error) {
+    console.error("[SceneFusion] Failed to download/upload image:", error);
+    throw error;
+  }
+}
+
+/**
+ * 调用阿里生图 API
+ */
+async function callAlibabaImageAPI(
+  baseURL: string,
+  apiKey: string,
+  modelId: string,
+  productImageUrl: string,
+  sceneImageUrl: string,
+  prompt: string
+): Promise<string> {
+  // 构建消息内容：产品图 + 场景图 + 提示词
+  const content = [
+    { image: productImageUrl },
+    { image: sceneImageUrl },
+    { text: prompt },
+  ];
+  
+  const requestBody = {
+    model: modelId,
+    input: {
+      messages: [
+        {
+          role: "user",
+          content,
+        },
+      ],
+    },
+    parameters: {
+      n: 1,
+      prompt_extend: true,
+      watermark: false,
+    },
+  };
+  
+  console.log("[SceneFusion] Sending request to Alibaba API...");
+  
+  const response = await fetch(baseURL, {
+    method: "POST",
+    headers: {
+      "Content-Type": "application/json",
+      Authorization: `Bearer ${apiKey}`,
+    },
+    body: JSON.stringify(requestBody),
+  });
+  
+  const data = await response.json();
+  
+  if (!response.ok || data.code) {
+    console.error("[SceneFusion] API Error:", data);
+    throw new Error(data.message || `API request failed: ${response.status}`);
+  }
+  
+  // 提取生成的图片 URL
+  const choices = data.output?.choices;
+  if (!choices || choices.length === 0) {
+    throw new Error("No images in response");
+  }
+  
+  const imageContent = choices[0]?.message?.content;
+  if (!imageContent || imageContent.length === 0) {
+    throw new Error("No image content in response");
+  }
+  
+  const generatedImageUrl = imageContent[0]?.image;
+  if (!generatedImageUrl) {
+    throw new Error("No image URL in response");
+  }
+  
+  console.log("[SceneFusion] Generated image URL:", generatedImageUrl.substring(0, 100) + "...");
+  return generatedImageUrl;
+}
+
+/**
+ * 处理单个场景
+ */
+async function processScene(
+  productImageUrl: string,
+  sceneImageUrl: string,
+  customPrompt: string | undefined,
+  defaultPrompt: string,
+  baseURL: string,
+  apiKey: string,
+  modelId: string
+): Promise<{ thumbnail: string; original: string }> {
+  // 组合提示词
+  const prompt = customPrompt 
+    ? `${defaultPrompt} ${customPrompt}` 
+    : defaultPrompt;
+  
+  // 调用阿里生图 API
+  const generatedImageUrl = await callAlibabaImageAPI(
+    baseURL,
+    apiKey,
+    modelId,
+    productImageUrl,
+    sceneImageUrl,
+    prompt
+  );
+  
+  // 下载并上传到 S3
+  const uploaded = await downloadAndUploadToS3(generatedImageUrl, "scene-fusion");
+  
+  return uploaded;
+}
+
+export async function POST(request: Request) {
+  try {
+    // 1. 认证
+    const session = await auth();
+    if (!session?.user?.id) {
+      return NextResponse.json(
+        { error: "未登录" },
+        { status: 401 }
+      );
+    }
+    
+    // 2. 获取场景裂变配置
+    const config = await getSceneFusionConfig();
+    
+    if (!config.enabled) {
+      return NextResponse.json(
+        { error: "场景裂变功能已被管理员禁用" },
+        { status: 403 }
+      );
+    }
+    
+    // 3. 解析请求体
+    const json = await request.json();
+    const parseResult = sceneFusionRequestSchema.safeParse(json);
+    
+    if (!parseResult.success) {
+      return NextResponse.json(
+        { error: parseResult.error.errors[0].message },
+        { status: 400 }
+      );
+    }
+    
+    const { productImage, scenes } = parseResult.data;
+    
+    // 检查场景数量限制
+    if (scenes.length > config.maxScenes) {
+      return NextResponse.json(
+        { error: `最多只能添加 ${config.maxScenes} 个场景` },
+        { status: 400 }
+      );
+    }
+    
+    console.log(`[SceneFusion] Processing ${scenes.length} scenes for user ${session.user.id}`);
+    console.log(`[SceneFusion] Using prompt: ${config.defaultPrompt.substring(0, 50)}...`);
+    
+    // 4. 获取阿里生图模型配置
+    const allModels = await getEnabledAiModelsWithProvider();
+    const alibabaModel = allModels.find(
+      (m) => m.provider.type === "alibaba-image" && m.model.isEnabled
+    );
+    
+    if (!alibabaModel) {
+      return NextResponse.json(
+        { error: "未配置阿里生图模型，请联系管理员" },
+        { status: 500 }
+      );
+    }
+    
+    const { provider, model } = alibabaModel;
+    console.log(`[SceneFusion] Using model: ${model.displayName} (${model.modelId})`);
+    
+    // 4. 预处理：将 Base64 图片上传到 S3
+    let productImageUrl = productImage;
+    if (productImage.startsWith("data:")) {
+      productImageUrl = await uploadBase64ToS3(productImage, "scene-fusion-product");
+    }
+    
+    const processedScenes = await Promise.all(
+      scenes.map(async (scene) => {
+        let sceneImageUrl = scene.image;
+        if (scene.image.startsWith("data:")) {
+          sceneImageUrl = await uploadBase64ToS3(scene.image, "scene-fusion-scene");
+        }
+        return {
+          ...scene,
+          imageUrl: sceneImageUrl,
+        };
+      })
+    );
+    
+    // 6. 并发处理所有场景
+    const results = await Promise.all(
+      processedScenes.map(async (scene) => {
+        try {
+          const result = await processScene(
+            productImageUrl,
+            scene.imageUrl,
+            scene.customPrompt,
+            config.defaultPrompt,
+            provider.baseUrl,
+            provider.apiKey,
+            model.modelId
+          );
+          
+          // 保存到数据库
+          const messageId = `scene-fusion-${Date.now()}-${scene.id}`;
+          const chatId = `scene-fusion-batch-${Date.now()}`;
+          
+          try {
+            await db.insert(generatedImage).values({
+              userId: session.user.id,
+              chatId,
+              messageId,
+              thumbnailUrl: result.thumbnail,
+              originalUrl: result.original,
+              imageSize: "2K",
+              prompt: scene.customPrompt 
+                ? `${config.defaultPrompt} ${scene.customPrompt}` 
+                : config.defaultPrompt,
+              modelId: model.modelId,
+            });
+            console.log(`[SceneFusion] Saved image to gallery: ${result.original.substring(0, 80)}...`);
+          } catch (dbError) {
+            console.error("[SceneFusion] Failed to save to gallery:", dbError);
+          }
+          
+          return {
+            id: scene.id,
+            status: "completed" as const,
+            result: result.thumbnail, // 返回缩略图用于预览
+            originalUrl: result.original, // 原图用于下载
+          };
+        } catch (error) {
+          console.error(`[SceneFusion] Failed to process scene ${scene.id}:`, error);
+          return {
+            id: scene.id,
+            status: "failed" as const,
+            error: error instanceof Error ? error.message : "处理失败",
+          };
+        }
+      })
+    );
+    
+    // 6. 返回结果
+    const successCount = results.filter((r) => r.status === "completed").length;
+    const failCount = results.filter((r) => r.status === "failed").length;
+    
+    console.log(`[SceneFusion] Completed: ${successCount} success, ${failCount} failed`);
+    
+    return NextResponse.json({
+      success: true,
+      results,
+      summary: {
+        total: scenes.length,
+        success: successCount,
+        failed: failCount,
+      },
+    });
+  } catch (error) {
+    console.error("[SceneFusion] Unexpected error:", error);
+    return NextResponse.json(
+      { error: error instanceof Error ? error.message : "处理失败" },
+      { status: 500 }
+    );
+  }
+}
diff --git a/components/image-annotator.tsx b/components/image-annotator.tsx
index 91349adf0..a44aa3b7a 100644
--- a/components/image-annotator.tsx
+++ b/components/image-annotator.tsx
@@ -4,11 +4,12 @@ import React, { useRef, useState, useEffect, useCallback } from "react";
 import { createPortal } from "react-dom";
 import Link from "next/link";
 import { useTheme } from "next-themes";
-import { X, Pencil, Circle, ArrowRight, Type, Check, Undo, Trash2, Scissors, Sparkles, Download, Expand, Move, Maximize2, ZoomIn, ZoomOut, Loader2, Move3D } from "lucide-react";
+import { X, Pencil, Circle, ArrowRight, Type, Check, Undo, Trash2, Scissors, Sparkles, Download, Expand, Move, Maximize2, ZoomIn, ZoomOut, Loader2, Move3D, Layers, Upload as UploadIcon, Image as ImageIcon } from "lucide-react";
 import { Button } from "./ui/button";
 import { Input } from "./ui/input";
 import { toast } from "sonner";
 import { AngleEditDialog } from "./angle-edit-dialog";
+import { SceneFusionDialog } from "./scene-fusion-dialog";
 
 type Tool = "pen" | "arrow" | "circle" | "text";
 
@@ -27,7 +28,7 @@ type DrawingAction = {
 };
 
 interface ImageAnnotatorProps {
-  imageUrl: string;
+  imageUrl?: string;
   onConfirm: (annotatedImageUrl: string) => void;
   onCancel: () => void;
   initialResizeMode?: boolean; // 是否直接进入尺寸调整模式
@@ -62,17 +63,101 @@ export function ImageAnnotator({ imageUrl, onConfirm, onCancel, initialResizeMod
   const [textPosition, setTextPosition] = useState<{ x: number; y: number } | null>(null);
   const [scale, setScale] = useState(1);
   const [offset, setOffset] = useState({ x: 0, y: 0 });
-  const [isLoading, setIsLoading] = useState(true);
+  const [isLoading, setIsLoading] = useState(!!imageUrl); // 如果有 imageUrl 则加载中
   const [isRemovingBg, setIsRemovingBg] = useState(false);
   const [isEnhancing, setIsEnhancing] = useState(false);
   const [isExtending, setIsExtending] = useState(false);
   const [showAngleEditDialog, setShowAngleEditDialog] = useState(false);
-  const [originalImageUrl, setOriginalImageUrl] = useState<string>(imageUrl);
+  const [showSceneFusion, setShowSceneFusion] = useState(false);
+  const [originalImageUrl, setOriginalImageUrl] = useState<string | undefined>(imageUrl);
   const [processedImageUrl, setProcessedImageUrl] = useState<string | null>(null);
   const [processedImageBase64, setProcessedImageBase64] = useState<string | null>(null);
   const [processType, setProcessType] = useState<"rmbg" | "enhance" | "extend" | "resize" | null>(null);
   const [mounted, setMounted] = useState(false);
   
+  const fileInputRef = useRef<HTMLInputElement>(null);
+
+  // 处理拖拽上传
+  const handleDrop = useCallback((e: React.DragEvent<HTMLDivElement>) => {
+    e.preventDefault();
+    e.stopPropagation();
+    
+    const files = e.dataTransfer.files;
+    if (files.length > 0) {
+      const file = files[0];
+      if (!file.type.startsWith('image/')) {
+        toast.error("请上传图片文件");
+        return;
+      }
+      if (file.size > 10 * 1024 * 1024) {
+        toast.error("图片大小不能超过 10MB");
+        return;
+      }
+      const reader = new FileReader();
+      reader.onload = (e) => {
+        const result = e.target?.result as string;
+        setOriginalImageUrl(result);
+        setIsLoading(true);
+        setProcessedImageUrl(null);
+        setProcessedImageBase64(null);
+        setProcessType(null);
+        setActions([]);
+        setCurrentAction(null);
+        
+        const img = new window.Image();
+        img.src = result;
+      };
+      reader.readAsDataURL(file);
+    }
+  }, []);
+
+  const handleDragOver = useCallback((e: React.DragEvent<HTMLDivElement>) => {
+    e.preventDefault();
+    e.stopPropagation();
+  }, []);
+
+  // 处理粘贴上传
+  useEffect(() => {
+    if (!mounted || originalImageUrl) return;
+    
+    const handlePaste = (e: ClipboardEvent) => {
+      const items = e.clipboardData?.items;
+      if (!items) return;
+      
+      for (const item of items) {
+        if (item.type.startsWith('image/')) {
+          const file = item.getAsFile();
+          if (!file) continue;
+          
+          if (file.size > 10 * 1024 * 1024) {
+            toast.error("图片大小不能超过 10MB");
+            return;
+          }
+          
+          const reader = new FileReader();
+          reader.onload = (e) => {
+            const result = e.target?.result as string;
+            setOriginalImageUrl(result);
+            setIsLoading(true);
+            setProcessedImageUrl(null);
+            setProcessedImageBase64(null);
+            setProcessType(null);
+            setActions([]);
+            setCurrentAction(null);
+            
+            const img = new window.Image();
+            img.src = result;
+          };
+          reader.readAsDataURL(file);
+          break;
+        }
+      }
+    };
+
+    document.addEventListener('paste', handlePaste);
+    return () => document.removeEventListener('paste', handlePaste);
+  }, [mounted, originalImageUrl]);
+
   // 尺寸调整相关状态
   const [isResizeMode, setIsResizeMode] = useState(false);
   const [resizeCanvasWidth, setResizeCanvasWidth] = useState<number>(1920);
@@ -169,8 +254,45 @@ export function ImageAnnotator({ imageUrl, onConfirm, onCancel, initialResizeMod
       });
   }, []);
 
+  // 处理本地图片上传
+  const handleImageUpload = (e: React.ChangeEvent<HTMLInputElement>) => {
+    const file = e.target.files?.[0];
+    if (file) {
+      if (file.size > 10 * 1024 * 1024) {
+        toast.error("图片大小不能超过 10MB");
+        return;
+      }
+      const reader = new FileReader();
+      reader.onload = (e) => {
+        const result = e.target?.result as string;
+        setOriginalImageUrl(result);
+        setIsLoading(true);
+        // 重置状态
+        setProcessedImageUrl(null);
+        setProcessedImageBase64(null);
+        setProcessType(null);
+        setActions([]);
+        setCurrentAction(null);
+        
+        // 触发图片加载逻辑
+        const img = new window.Image();
+        img.crossOrigin = "anonymous";
+        img.onload = () => {
+          setImage(img);
+          setIsLoading(false);
+        };
+        img.src = result;
+      };
+      reader.readAsDataURL(file);
+    }
+    // 重置 input
+    e.target.value = "";
+  };
+
   // 加载图片
   useEffect(() => {
+    if (!imageUrl) return; // 如果没有初始图片，不执行加载逻辑
+
     setIsLoading(true);
     setOriginalImageUrl(imageUrl);
     setProcessedImageUrl(null);
@@ -459,7 +581,7 @@ export function ImageAnnotator({ imageUrl, onConfirm, onCancel, initialResizeMod
 
   // 抠图功能
   const handleRemoveBackground = async () => {
-    if (!image) return;
+    if (!image || !originalImageUrl) return;
 
     setIsRemovingBg(true);
     try {
@@ -522,7 +644,7 @@ export function ImageAnnotator({ imageUrl, onConfirm, onCancel, initialResizeMod
 
   // 变清晰处理
   const handleEnhance = async () => {
-    if (!image) return;
+    if (!image || !originalImageUrl) return;
 
     setIsEnhancing(true);
     try {
@@ -1170,6 +1292,16 @@ export function ImageAnnotator({ imageUrl, onConfirm, onCancel, initialResizeMod
               <Move3D className="mr-1 h-4 w-4" />
               角度调整
             </Button>
+            <Button 
+              variant="ghost" 
+              size="sm" 
+              onClick={() => setShowSceneFusion(true)} 
+              disabled={showSceneFusion || showAngleEditDialog || isRemovingBg || isEnhancing || isExtending || isResizeMode || !image}
+              title="场景裂变"
+            >
+              <Layers className="mr-1 h-4 w-4" />
+              场景裂变
+            </Button>
             <Button variant="ghost" size="sm" onClick={handleUndo} disabled={actions.length === 0}>
               <Undo className="mr-1 h-4 w-4" />
               撤销
@@ -1574,7 +1706,7 @@ export function ImageAnnotator({ imageUrl, onConfirm, onCancel, initialResizeMod
       )}
 
       {/* 角度调整对话框 */}
-      {showAngleEditDialog && (
+      {showAngleEditDialog && originalImageUrl && (
         <AngleEditDialog
           open={showAngleEditDialog}
           onOpenChange={setShowAngleEditDialog}
@@ -1582,6 +1714,12 @@ export function ImageAnnotator({ imageUrl, onConfirm, onCancel, initialResizeMod
           onComplete={handleAngleEditComplete}
         />
       )}
+
+      {/* 场景裂变弹窗 */}
+      <SceneFusionDialog 
+        open={showSceneFusion} 
+        onOpenChange={setShowSceneFusion} 
+      />
     </div>
   );
 
@@ -1590,5 +1728,55 @@ export function ImageAnnotator({ imageUrl, onConfirm, onCancel, initialResizeMod
     return null;
   }
 
+  // 如果没有图片，显示上传界面
+  if (!originalImageUrl) {
+    return createPortal(
+      <div className="fixed inset-0 z-50 flex items-center justify-center bg-black/80 backdrop-blur-sm">
+        <div className="relative flex h-[80vh] w-[90vw] max-w-5xl flex-col overflow-hidden rounded-xl border bg-background shadow-2xl animate-in fade-in zoom-in-95 duration-200">
+          {/* Header */}
+          <div className="flex h-14 items-center justify-between border-b px-4">
+            <div className="flex items-center gap-2">
+              <span className="text-lg font-bold">图片编辑器</span>
+            </div>
+            <Button
+              variant="ghost"
+              size="icon"
+              onClick={onCancel}
+              className="h-8 w-8 rounded-full"
+            >
+              <X className="h-4 w-4" />
+            </Button>
+          </div>
+
+          {/* Upload Area */}
+          <div className="flex flex-1 flex-col items-center justify-center p-8 bg-muted/10">
+            <div 
+              className="w-full max-w-md border-2 border-dashed border-muted-foreground/25 rounded-xl p-12 text-center cursor-pointer hover:border-primary/50 hover:bg-muted/20 transition-all group"
+              onClick={() => fileInputRef.current?.click()}
+              onDrop={handleDrop}
+              onDragOver={handleDragOver}
+            >
+              <div className="bg-background rounded-full p-4 w-20 h-20 mx-auto mb-6 shadow-sm group-hover:scale-105 transition-transform flex items-center justify-center">
+                <UploadIcon className="w-10 h-10 text-muted-foreground group-hover:text-primary transition-colors" />
+              </div>
+              <h3 className="text-xl font-medium mb-2">上传图片以开始编辑</h3>
+              <p className="text-muted-foreground mb-2">拖拽图片到此处、点击选择或 Ctrl+V 粘贴</p>
+              <p className="text-muted-foreground/60 text-sm mb-6">支持 JPG, PNG, WebP 格式，最大 10MB</p>
+              <Button variant="default">选择图片</Button>
+              <input 
+                type="file" 
+                ref={fileInputRef}
+                className="hidden" 
+                accept="image/png,image/jpeg,image/webp"
+                onChange={handleImageUpload}
+              />
+            </div>
+          </div>
+        </div>
+      </div>,
+      document.body
+    );
+  }
+
   return createPortal(modalContent, document.body);
 }
diff --git a/components/scene-fusion-dialog.tsx b/components/scene-fusion-dialog.tsx
new file mode 100644
index 000000000..e465db802
--- /dev/null
+++ b/components/scene-fusion-dialog.tsx
@@ -0,0 +1,568 @@
+"use client";
+
+import React, { useState, useRef } from "react";
+import useSWR from "swr";
+import { 
+  Dialog, 
+  DialogContent, 
+  DialogHeader, 
+  DialogTitle, 
+  DialogTrigger,
+  DialogFooter,
+  DialogDescription
+} from "@/components/ui/dialog";
+import { Button } from "@/components/ui/button";
+import { Input } from "@/components/ui/input";
+import { Label } from "@/components/ui/label";
+import { ScrollArea } from "@/components/ui/scroll-area";
+import { 
+  Upload, 
+  Plus, 
+  Trash2, 
+  Download, 
+  RefreshCw, 
+  Image as ImageIcon, 
+  Loader2, 
+  X,
+  Layers,
+  Sparkles
+} from "lucide-react";
+import { cn } from "@/lib/utils";
+import { toast } from "sonner";
+
+const fetcher = (url: string) => fetch(url).then((res) => res.json());
+
+// 场景项数据结构
+export interface SceneItem {
+  id: string;
+  image: string; // 场景图 URL 或 Base64
+  customPrompt: string; // 该场景的补充关键词
+  status: 'pending' | 'processing' | 'completed' | 'failed';
+  result?: string; // 生成结果缩略图 URL
+  originalUrl?: string; // 生成结果原图 URL（用于下载）
+  error?: string; // 错误信息
+}
+
+interface SceneFusionDialogProps {
+  open: boolean;
+  onOpenChange: (open: boolean) => void;
+}
+
+export function SceneFusionDialog({ open, onOpenChange }: SceneFusionDialogProps) {
+  // 获取场景裂变配置（使用公开 API 端点）
+  const { data: configData } = useSWR(
+    open ? "/api/scene-fusion/config" : null,
+    fetcher
+  );
+  
+  const config = configData as {
+    enabled?: boolean;
+    defaultPrompt?: string;
+    maxScenes?: number;
+  } | undefined;
+  
+  const defaultPrompt = config?.defaultPrompt || "将图1的产品替换到图2的场景中，保持图1产品的外观、颜色和细节不变，自然融合到图2的环境中。";
+  const maxScenes = config?.maxScenes || 10;
+  
+  // 状态管理
+  const [productImage, setProductImage] = useState<string | null>(null);
+  const [scenes, setScenes] = useState<SceneItem[]>([]);
+  const [isProcessing, setIsProcessing] = useState(false);
+  
+  // 文件上传 Input 引用
+  const productInputRef = useRef<HTMLInputElement>(null);
+  const sceneInputRef = useRef<HTMLInputElement>(null);
+
+  // 处理产品图上传
+  const handleProductUpload = (e: React.ChangeEvent<HTMLInputElement>) => {
+    const file = e.target.files?.[0];
+    if (!file) return;
+
+    if (file.size > 10 * 1024 * 1024) {
+      toast.error("图片大小不能超过 10MB");
+      return;
+    }
+
+    const reader = new FileReader();
+    reader.onload = (e) => {
+      setProductImage(e.target?.result as string);
+    };
+    reader.readAsDataURL(file);
+    // 重置 input，允许重复选择同一文件
+    e.target.value = "";
+  };
+
+  // 处理场景图上传
+  const handleSceneUpload = (e: React.ChangeEvent<HTMLInputElement>) => {
+    const files = Array.from(e.target.files || []);
+    if (files.length === 0) return;
+
+    if (scenes.length + files.length > maxScenes) {
+      toast.error(`最多只能添加 ${maxScenes} 张场景参考图`);
+      return;
+    }
+
+    files.forEach(file => {
+      if (file.size > 10 * 1024 * 1024) {
+        toast.error(`图片 ${file.name} 大小不能超过 10MB`);
+        return;
+      }
+
+      const reader = new FileReader();
+      reader.onload = (e) => {
+        const newScene: SceneItem = {
+          id: Math.random().toString(36).substring(7),
+          image: e.target?.result as string,
+          customPrompt: "",
+          status: 'pending'
+        };
+        setScenes(prev => [...prev, newScene]);
+      };
+      reader.readAsDataURL(file);
+    });
+    
+    // 重置 input
+    e.target.value = "";
+  };
+
+  // 移除场景
+  const removeScene = (id: string) => {
+    setScenes(prev => prev.filter(scene => scene.id !== id));
+  };
+
+  // 更新场景关键词
+  const updateScenePrompt = (id: string, prompt: string) => {
+    setScenes(prev => prev.map(scene => 
+      scene.id === id ? { ...scene, customPrompt: prompt } : scene
+    ));
+  };
+
+  // 开始裂变
+  const startFusion = async () => {
+    if (config?.enabled === false) {
+      toast.error("场景裂变功能已被管理员禁用");
+      return;
+    }
+    if (!productImage) {
+      toast.error("请先上传产品图");
+      return;
+    }
+    if (scenes.length === 0) {
+      toast.error("请至少添加一张场景参考图");
+      return;
+    }
+
+    setIsProcessing(true);
+    
+    // 将所有 pending 或 failed 的状态改为 processing
+    const scenesToProcess = scenes.filter(s => s.status === 'pending' || s.status === 'failed');
+    setScenes(prev => prev.map(s => 
+      (s.status === 'pending' || s.status === 'failed') 
+        ? { ...s, status: 'processing' } 
+        : s
+    ));
+
+    try {
+      // 调用后端 API
+      const response = await fetch('/api/scene-fusion', {
+        method: 'POST',
+        headers: {
+          'Content-Type': 'application/json',
+        },
+        body: JSON.stringify({
+          productImage,
+          scenes: scenesToProcess.map(s => ({
+            id: s.id,
+            image: s.image,
+            customPrompt: s.customPrompt || undefined,
+          })),
+        }),
+      });
+
+      const data = await response.json();
+
+      if (!response.ok) {
+        throw new Error(data.error || '处理失败');
+      }
+
+      // 更新场景状态
+      setScenes(prev => prev.map(s => {
+        const result = data.results?.find((r: any) => r.id === s.id);
+        if (result) {
+          if (result.status === 'completed') {
+            return { 
+              ...s, 
+              status: 'completed', 
+              result: result.result,
+              originalUrl: result.originalUrl,
+            };
+          } else {
+            return { 
+              ...s, 
+              status: 'failed', 
+              error: result.error || '生成失败' 
+            };
+          }
+        }
+        return s;
+      }));
+
+      const { summary } = data;
+      if (summary.failed > 0) {
+        toast.warning(`批量处理完成：${summary.success} 成功，${summary.failed} 失败`);
+      } else {
+        toast.success(`批量处理完成：${summary.success} 张图片已生成`);
+      }
+    } catch (error) {
+      console.error('[SceneFusion] Error:', error);
+      toast.error(error instanceof Error ? error.message : '处理失败');
+      
+      // 将 processing 状态改回 failed
+      setScenes(prev => prev.map(s => 
+        s.status === 'processing' ? { ...s, status: 'failed', error: '处理失败' } : s
+      ));
+    } finally {
+      setIsProcessing(false);
+    }
+  };
+
+  // 重新生成单个场景
+  const regenerateScene = async (id: string) => {
+    if (!productImage) return;
+    
+    const scene = scenes.find(s => s.id === id);
+    if (!scene) return;
+
+    setScenes(prev => prev.map(s => 
+      s.id === id ? { ...s, status: 'processing', error: undefined } : s
+    ));
+
+    try {
+      const response = await fetch('/api/scene-fusion', {
+        method: 'POST',
+        headers: {
+          'Content-Type': 'application/json',
+        },
+        body: JSON.stringify({
+          productImage,
+          scenes: [{
+            id: scene.id,
+            image: scene.image,
+            customPrompt: scene.customPrompt || undefined,
+          }],
+        }),
+      });
+
+      const data = await response.json();
+
+      if (!response.ok) {
+        throw new Error(data.error || '处理失败');
+      }
+
+      const result = data.results?.[0];
+      if (result?.status === 'completed') {
+        setScenes(prev => prev.map(s => 
+          s.id === id ? { 
+            ...s, 
+            status: 'completed', 
+            result: result.result,
+            originalUrl: result.originalUrl,
+          } : s
+        ));
+        toast.success('重新生成成功');
+      } else {
+        throw new Error(result?.error || '生成失败');
+      }
+    } catch (error) {
+      console.error('[SceneFusion] Regenerate error:', error);
+      setScenes(prev => prev.map(s => 
+        s.id === id ? { ...s, status: 'failed', error: error instanceof Error ? error.message : '处理失败' } : s
+      ));
+      toast.error(error instanceof Error ? error.message : '重新生成失败');
+    }
+  };
+
+  // 下载图片
+  const downloadImage = (url: string, filename: string) => {
+    const link = document.createElement('a');
+    link.href = url;
+    link.download = filename;
+    document.body.appendChild(link);
+    link.click();
+    document.body.removeChild(link);
+  };
+
+  // 全部下载
+  const downloadAll = async () => {
+    const completedScenes = scenes.filter(s => s.status === 'completed' && (s.originalUrl || s.result));
+    if (completedScenes.length === 0) {
+      toast.error("没有可下载的图片");
+      return;
+    }
+
+    toast.info(`正在下载 ${completedScenes.length} 张图片...`);
+    
+    for (let i = 0; i < completedScenes.length; i++) {
+      const scene = completedScenes[i];
+      const url = scene.originalUrl || scene.result!;
+      // 延迟下载避免浏览器阻止
+      setTimeout(() => {
+        downloadImage(url, `scene-fusion-${i + 1}.png`);
+      }, i * 300);
+    }
+  };
+
+  return (
+    <Dialog open={open} onOpenChange={onOpenChange}>
+      <DialogContent className="max-w-5xl h-[90vh] flex flex-col p-0 gap-0">
+        <DialogHeader className="px-6 py-4 border-b shrink-0">
+          <DialogTitle className="flex items-center gap-2">
+            <Layers className="w-5 h-5 text-primary" />
+            场景裂变
+          </DialogTitle>
+          <DialogDescription>
+            上传1张产品图和多张场景图，AI 自动将产品融合到所有场景中。
+          </DialogDescription>
+        </DialogHeader>
+
+        <div className="flex-1 overflow-hidden flex flex-col md:flex-row">
+          {/* 左侧：输入区域 */}
+          <div className="w-full md:w-[400px] border-r flex flex-col bg-muted/10 shrink-0">
+            <ScrollArea className="flex-1 p-4">
+              <div className="space-y-6">
+                {/* 1. 产品图上传 */}
+                <div className="space-y-2">
+                  <div className="flex items-center justify-between">
+                    <Label className="text-base font-medium">产品图 A (必须)</Label>
+                    {productImage && (
+                      <Button variant="ghost" size="sm" onClick={() => setProductImage(null)} className="h-6 text-xs text-muted-foreground hover:text-destructive">
+                        清除
+                      </Button>
+                    )}
+                  </div>
+                  
+                  <div 
+                    className={cn(
+                      "border-2 border-dashed rounded-lg p-4 transition-colors text-center cursor-pointer relative group",
+                      productImage ? "border-primary/50 bg-primary/5" : "border-muted-foreground/25 hover:border-primary/50 hover:bg-muted/50"
+                    )}
+                    onClick={() => !productImage && productInputRef.current?.click()}
+                  >
+                    <input 
+                      type="file" 
+                      ref={productInputRef}
+                      className="hidden" 
+                      accept="image/png,image/jpeg,image/webp"
+                      onChange={handleProductUpload}
+                    />
+                    
+                    {productImage ? (
+                      <div className="relative aspect-square w-full max-w-[200px] mx-auto">
+                        <img src={productImage} alt="产品图" className="w-full h-full object-contain rounded-md" />
+                        <div className="absolute inset-0 bg-black/40 opacity-0 group-hover:opacity-100 transition-opacity flex items-center justify-center rounded-md">
+                          <Button variant="secondary" size="sm" onClick={(e) => { e.stopPropagation(); productInputRef.current?.click(); }}>
+                            更换图片
+                          </Button>
+                        </div>
+                      </div>
+                    ) : (
+                      <div className="py-8 flex flex-col items-center gap-2 text-muted-foreground">
+                        <Upload className="w-8 h-8 opacity-50" />
+                        <span className="text-sm">点击上传产品图</span>
+                      </div>
+                    )}
+                  </div>
+                </div>
+
+                {/* 2. 场景图列表 */}
+                <div className="space-y-2">
+                  <div className="flex items-center justify-between">
+                    <Label className="text-base font-medium">
+                      场景参考图 ({scenes.length}/{maxScenes})
+                    </Label>
+                    <Button 
+                      variant="outline" 
+                      size="sm" 
+                      onClick={() => sceneInputRef.current?.click()}
+                      disabled={scenes.length >= maxScenes || isProcessing}
+                      className="h-7 text-xs gap-1"
+                    >
+                      <Plus className="w-3 h-3" /> 添加场景
+                    </Button>
+                    <input 
+                      type="file" 
+                      ref={sceneInputRef}
+                      className="hidden" 
+                      multiple
+                      accept="image/png,image/jpeg,image/webp"
+                      onChange={handleSceneUpload}
+                    />
+                  </div>
+
+                  <div className="space-y-3">
+                    {scenes.length === 0 && (
+                      <div 
+                        className="border-2 border-dashed border-muted-foreground/25 rounded-lg p-8 text-center cursor-pointer hover:bg-muted/50 transition-colors"
+                        onClick={() => sceneInputRef.current?.click()}
+                      >
+                        <ImageIcon className="w-8 h-8 mx-auto opacity-20 mb-2" />
+                        <p className="text-sm text-muted-foreground">点击添加场景参考图</p>
+                        <p className="text-xs text-muted-foreground/60 mt-1">支持批量上传，最多{maxScenes}张</p>
+                      </div>
+                    )}
+
+                    {scenes.map((scene, index) => (
+                      <div key={scene.id} className="border rounded-lg p-3 bg-background relative group">
+                        <div className="flex gap-3">
+                          {/* 缩略图 */}
+                          <div className="w-20 h-20 shrink-0 bg-muted rounded-md overflow-hidden border">
+                            <img src={scene.image} alt={`场景 ${index + 1}`} className="w-full h-full object-cover" />
+                          </div>
+                          
+                          {/* 关键词输入 */}
+                          <div className="flex-1 space-y-1.5">
+                            <div className="flex items-center justify-between">
+                              <span className="text-xs font-medium text-muted-foreground">场景 {index + 1}</span>
+                              {!isProcessing && (
+                                <Button 
+                                  variant="ghost" 
+                                  size="icon" 
+                                  className="h-6 w-6 text-muted-foreground hover:text-destructive"
+                                  onClick={() => removeScene(scene.id)}
+                                >
+                                  <Trash2 className="w-3.5 h-3.5" />
+                                </Button>
+                              )}
+                            </div>
+                            <Input 
+                              placeholder="补充关键词：如'放在木桌上'" 
+                              value={scene.customPrompt}
+                              onChange={(e) => updateScenePrompt(scene.id, e.target.value)}
+                              className="h-8 text-xs"
+                              disabled={isProcessing}
+                            />
+                          </div>
+                        </div>
+                      </div>
+                    ))}
+                  </div>
+                </div>
+              </div>
+            </ScrollArea>
+            
+            {/* 默认提示词提示 */}
+            <div className="px-4 py-2 border-t bg-muted/30">
+              <p className="text-xs text-muted-foreground">
+                <span className="font-medium">默认提示词：</span>
+                {defaultPrompt.length > 60 ? `${defaultPrompt.substring(0, 60)}...` : defaultPrompt}
+              </p>
+            </div>
+            
+            <div className="p-4 border-t bg-background">
+              <Button 
+                className="w-full" 
+                size="lg" 
+                onClick={startFusion}
+                disabled={isProcessing || !productImage || scenes.length === 0}
+              >
+                {isProcessing ? (
+                  <>
+                    <Loader2 className="w-4 h-4 mr-2 animate-spin" />
+                    处理中...
+                  </>
+                ) : (
+                  <>
+                    <Sparkles className="w-4 h-4 mr-2" />
+                    开始裂变 ({scenes.length} 张)
+                  </>
+                )}
+              </Button>
+            </div>
+          </div>
+
+          {/* 右侧：结果展示区域 */}
+          <div className="flex-1 flex flex-col bg-background/50 h-full">
+            <div className="p-4 border-b flex items-center justify-between bg-background">
+              <h3 className="font-medium">生成结果</h3>
+              {scenes.some(s => s.status === 'completed') && (
+                <Button variant="outline" size="sm" onClick={downloadAll}>
+                  <Download className="w-4 h-4 mr-2" />
+                  全部下载
+                </Button>
+              )}
+            </div>
+            
+            <ScrollArea className="flex-1 p-4">
+              {scenes.length === 0 ? (
+                <div className="h-full flex flex-col items-center justify-center text-muted-foreground opacity-50 min-h-[300px]">
+                  <Layers className="w-16 h-16 mb-4" />
+                  <p>添加场景图并开始裂变，结果将显示在这里</p>
+                </div>
+              ) : (
+                <div className="grid grid-cols-1 md:grid-cols-2 lg:grid-cols-3 gap-4">
+                  {scenes.map((scene, index) => (
+                    <div key={scene.id} className="border rounded-lg overflow-hidden bg-background shadow-sm flex flex-col">
+                      <div className="p-2 border-b text-xs text-muted-foreground bg-muted/30 flex justify-between items-center">
+                        <span>场景 {index + 1}</span>
+                        <span className={cn(
+                          "capitalize px-1.5 py-0.5 rounded text-[10px]",
+                          scene.status === 'completed' && "bg-green-100 text-green-700 dark:bg-green-900/30 dark:text-green-400",
+                          scene.status === 'failed' && "bg-red-100 text-red-700 dark:bg-red-900/30 dark:text-red-400",
+                          scene.status === 'processing' && "bg-blue-100 text-blue-700 dark:bg-blue-900/30 dark:text-blue-400",
+                          scene.status === 'pending' && "bg-gray-100 text-gray-600 dark:bg-gray-800 dark:text-gray-400",
+                        )}>
+                          {scene.status === 'completed' ? '完成' : 
+                           scene.status === 'failed' ? '失败' : 
+                           scene.status === 'processing' ? '生成中' : '等待'}
+                        </span>
+                      </div>
+                      
+                      <div className="aspect-[4/3] bg-muted/20 relative group">
+                        {scene.result ? (
+                          <img src={scene.result} alt="生成结果" className="w-full h-full object-cover" />
+                        ) : (
+                          <div className="w-full h-full flex flex-col items-center justify-center p-4">
+                            {scene.status === 'processing' ? (
+                              <div className="flex flex-col items-center gap-2">
+                                <Loader2 className="w-8 h-8 animate-spin text-primary" />
+                                <span className="text-xs text-muted-foreground">正在生成...</span>
+                              </div>
+                            ) : scene.status === 'failed' ? (
+                              <div className="text-center">
+                                <p className="text-sm text-destructive mb-2">{scene.error || "生成失败"}</p>
+                                <Button size="sm" variant="outline" onClick={() => regenerateScene(scene.id)}>
+                                  <RefreshCw className="w-3 h-3 mr-1" /> 重试
+                                </Button>
+                              </div>
+                            ) : (
+                              <div className="flex items-center justify-center w-full h-full opacity-30">
+                                <img src={scene.image} className="w-full h-full object-cover blur-sm" />
+                                <span className="absolute text-xs font-medium bg-black/50 text-white px-2 py-1 rounded">等待处理</span>
+                              </div>
+                            )}
+                          </div>
+                        )}
+
+                        {/* 悬停操作层 */}
+                        {scene.status === 'completed' && (
+                          <div className="absolute inset-0 bg-black/60 opacity-0 group-hover:opacity-100 transition-opacity flex items-center justify-center gap-2">
+                            <Button size="icon" variant="secondary" onClick={() => downloadImage(scene.originalUrl || scene.result!, `scene-${index+1}.png`)}>
+                              <Download className="w-4 h-4" />
+                            </Button>
+                            <Button size="icon" variant="secondary" onClick={() => regenerateScene(scene.id)}>
+                              <RefreshCw className="w-4 h-4" />
+                            </Button>
+                          </div>
+                        )}
+                      </div>
+                    </div>
+                  ))}
+                </div>
+              )}
+            </ScrollArea>
+          </div>
+        </div>
+      </DialogContent>
+    </Dialog>
+  );
+}
diff --git a/components/top-navbar.tsx b/components/top-navbar.tsx
index 340424c3a..357400fc3 100644
--- a/components/top-navbar.tsx
+++ b/components/top-navbar.tsx
@@ -49,7 +49,7 @@ export function TopNavbar({
   const [rankingOpen, setRankingOpen] = useState(false);
   const [showImageEditor, setShowImageEditor] = useState(false);
   const [editorImageUrl, setEditorImageUrl] = useState<string>("");
-  const fileInputRef = useRef<HTMLInputElement>(null);
+  // 移除 fileInputRef，不再需要外部上传
 
   useEffect(() => {
     setMounted(true);
@@ -90,45 +90,12 @@ export function TopNavbar({
   };
 
   // 决定显示哪个 LOGO
-  // 导航栏背景固定为黑色，理论上应该始终显示适合深色背景的 LOGO
-  // 但如果有专门配置了 "Dark Mode Logo" (通常用于深色背景)，我们优先使用它
-  // 如果没配置 Dark Mode Logo，则使用默认 Logo
-  // 注意：这里的背景是黑色的，所以通常需要一个浅色的 LOGO（即 Dark Mode 下使用的 LOGO）
-  // 如果用户配置逻辑是：siteLogo(Light Mode/白底用), siteLogoDark(Dark Mode/黑底用)
-  // 那么在黑色导航栏上，应该优先使用 siteLogoDark
   const displayLogo = (siteLogoDark && siteLogoDark.trim() !== "") ? siteLogoDark : siteLogo;
 
-  // 处理图片上传
-  const handleImageUpload = (e: React.ChangeEvent<HTMLInputElement>) => {
-    const file = e.target.files?.[0];
-    if (!file) return;
-
-    // 检查文件类型
-    if (!file.type.startsWith("image/")) {
-      return;
-    }
-
-    // 转换为 data URL
-    const reader = new FileReader();
-    reader.onloadend = () => {
-      const result = reader.result as string;
-      setEditorImageUrl(result);
-      setShowImageEditor(true);
-    };
-    reader.onerror = () => {
-      console.error("图片读取失败");
-    };
-    reader.readAsDataURL(file);
-
-    // 清空文件输入，以便可以再次选择同一文件
-    if (fileInputRef.current) {
-      fileInputRef.current.value = "";
-    }
-  };
-
-  // 打开图片编辑器
+  // 打开图片编辑器 - 直接打开，不触发上传
   const handleOpenImageEditor = () => {
-    fileInputRef.current?.click();
+    setEditorImageUrl(""); // 确保初始为空
+    setShowImageEditor(true);
   };
 
   // 处理图片编辑完成
@@ -287,15 +254,6 @@ export function TopNavbar({
           <span className="text-sm font-medium">图片编辑</span>
         </Button>
 
-        {/* 隐藏的文件输入 */}
-        <input
-          ref={fileInputRef}
-          type="file"
-          accept="image/*"
-          onChange={handleImageUpload}
-          className="hidden"
-        />
-
         {/* 我的图册 */}
         <Button
           variant="ghost"
@@ -330,7 +288,7 @@ export function TopNavbar({
       </div>
 
       {/* 图片编辑器 */}
-      {showImageEditor && editorImageUrl && (
+      {showImageEditor && (
         <ImageAnnotator
           imageUrl={editorImageUrl}
           onConfirm={handleImageEditorConfirm}
diff --git a/docs/图片编辑器功能升级规划.md b/docs/图片编辑器功能升级规划.md
new file mode 100644
index 000000000..7ef088f6c
--- /dev/null
+++ b/docs/图片编辑器功能升级规划.md
@@ -0,0 +1,303 @@
+# 图片编辑器功能升级规划文档
+
+## 概述
+
+本文档规划两个新功能：
+1. **图片编辑器入口优化** - 允许用户先打开页面再上传图片
+2. **场景裂变功能** - 批量将产品图替换到多个场景图中
+
+---
+
+## 功能一：图片编辑器入口优化
+
+### 当前问题
+- 用户点击顶部导航栏的「图片编辑器」按钮时，必须先上传图片才能打开界面
+- 用户体验不佳，无法预览编辑器界面
+
+### 解决方案
+
+#### 方案 A：两步式交互（推荐）
+1. 用户点击「图片编辑器」按钮
+2. 打开图片编辑器页面/弹窗
+3. 页面中显示上传区域，用户可以：
+   - 拖拽上传图片
+   - 点击选择图片
+   - 粘贴图片
+4. 上传后进入编辑模式
+
+#### UI 设计
+```
+┌─────────────────────────────────────────────────────────┐
+│                    图片编辑器                            │
+├─────────────────────────────────────────────────────────┤
+│                                                         │
+│     ┌─────────────────────────────────────────┐        │
+│     │                                         │        │
+│     │      📷 拖拽图片到此处或点击上传         │        │
+│     │                                         │        │
+│     │      支持 JPG、PNG、WebP 格式            │        │
+│     │      最大 10MB                          │        │
+│     │                                         │        │
+│     └─────────────────────────────────────────┘        │
+│                                                         │
+│     [选择图片]  [从剪贴板粘贴]                           │
+│                                                         │
+└─────────────────────────────────────────────────────────┘
+```
+
+#### 技术实现
+- 修改 `ImageAnnotator` 组件，支持无图片时显示上传界面
+- 修改导航栏点击事件，直接打开编辑器弹窗
+- 添加拖拽上传、点击上传、粘贴上传功能
+
+---
+
+## 功能二：场景裂变
+
+### 功能描述
+用户上传一张产品图（图A），再上传 1-10 张场景参考图（B1-B10），系统自动将产品图批量融合到每个场景中。
+
+### 用户流程
+```
+1. 用户上传产品图 A（必须）
+2. 用户上传场景图 B1-B10（至少1张，最多10张）
+3. 用户可选填补充关键词
+4. 点击「开始裂变」
+5. 系统依次调用阿里生图 API：
+   - 图A + 图B1 → 生成 AB1
+   - 图A + 图B2 → 生成 AB2
+   - ...
+   - 图A + 图B10 → 生成 AB10
+6. 显示所有生成结果，支持下载
+```
+
+### UI 设计
+```
+┌─────────────────────────────────────────────────────────────────┐
+│                        场景裂变                                  │
+├─────────────────────────────────────────────────────────────────┤
+│                                                                 │
+│  ┌─────────────────────────────────────────────────────────┐   │
+│  │  产品图 A (必须上传)                                      │   │
+│  │  ┌─────────┐                                             │   │
+│  │  │         │  [上传产品图]  [更换]                        │   │
+│  │  │  产品   │                                             │   │
+│  │  │         │                                             │   │
+│  │  └─────────┘                                             │   │
+│  └─────────────────────────────────────────────────────────┘   │
+│                                                                 │
+│  ┌─────────────────────────────────────────────────────────┐   │
+│  │  场景参考图 (1-10张)                          [添加场景] │   │
+│  │                                                         │   │
+│  │  ┌─────────────────────────────────────────────────┐   │   │
+│  │  │ 场景 1                                    [删除] │   │   │
+│  │  │ ┌───────┐  补充关键词：                          │   │   │
+│  │  │ │  B1   │  ┌─────────────────────────────────┐  │   │   │
+│  │  │ │       │  │ 例如：放在木质桌面上            │  │   │   │
+│  │  │ └───────┘  └─────────────────────────────────┘  │   │   │
+│  │  └─────────────────────────────────────────────────┘   │   │
+│  │                                                         │   │
+│  │  ┌─────────────────────────────────────────────────┐   │   │
+│  │  │ 场景 2                                    [删除] │   │   │
+│  │  │ ┌───────┐  补充关键词：                          │   │   │
+│  │  │ │  B2   │  ┌─────────────────────────────────┐  │   │   │
+│  │  │ │       │  │ 例如：放在白色大理石台面上      │  │   │   │
+│  │  │ └───────┘  └─────────────────────────────────┘  │   │   │
+│  │  └─────────────────────────────────────────────────┘   │   │
+│  │                                                         │   │
+│  │  ... (更多场景)                                         │   │
+│  │                                                         │   │
+│  └─────────────────────────────────────────────────────────┘   │
+│                                                                 │
+│  默认提示词：将图1的产品替换到图2的场景中，保持图1产品的外观、   │
+│  颜色和细节不变，自然融合到图2的环境中。                         │
+│                                                                 │
+│                    [开始裂变] (已添加 2 个场景)                   │
+│                                                                 │
+├─────────────────────────────────────────────────────────────────┤
+│  生成结果：                                                      │
+│                                                                 │
+│  ┌─────────────────┐ ┌─────────────────┐                       │
+│  │ 场景1: AB1      │ │ 场景2: AB2      │                       │
+│  │ ┌─────────┐     │ │ ┌─────────┐     │                       │
+│  │ │         │     │ │ │         │     │                       │
+│  │ │ ✓完成   │     │ │ │ 生成中..│     │                       │
+│  │ └─────────┘     │ │ └─────────┘     │                       │
+│  │ [下载] [重新生成]│ │ [下载] [重新生成]│                       │
+│  └─────────────────┘ └─────────────────┘                       │
+│                                                                 │
+│  [全部下载] [保存到相册]                                         │
+│                                                                 │
+└─────────────────────────────────────────────────────────────────┘
+```
+
+### 技术实现
+
+#### 1. 数据结构
+```typescript
+// 场景项数据结构
+interface SceneItem {
+  id: string;                // 唯一标识
+  image: string;             // 场景图 URL 或 Base64
+  customPrompt: string;      // 该场景的补充关键词
+  status: 'pending' | 'processing' | 'completed' | 'failed';
+  result?: string;           // 生成结果图片 URL
+  error?: string;            // 错误信息
+}
+
+// 场景裂变状态
+interface SceneFusionState {
+  productImage: string | null;  // 产品图 A
+  scenes: SceneItem[];          // 场景列表 (1-10)
+  isProcessing: boolean;        // 是否正在处理
+  currentIndex: number;         // 当前处理的场景索引
+}
+```
+
+#### 2. 新建组件
+- `components/scene-fusion-dialog.tsx` - 场景裂变弹窗组件
+
+#### 3. API 调用逻辑
+```typescript
+// 调用现有的阿里生图 API
+// 每次请求：产品图A + 场景图Bn + 提示词
+
+const defaultPrompt = "将图1的产品替换到图2的场景中，保持图1产品的外观、颜色和细节不变，自然融合到图2的环境中。";
+
+async function generateSceneFusion(
+  productImage: string,      // 产品图 A
+  sceneImage: string,        // 场景图 Bn
+  customPrompt?: string      // 该场景的补充关键词（每个场景独立）
+) {
+  // 组合提示词：默认提示词 + 该场景的补充关键词
+  const prompt = customPrompt 
+    ? `${defaultPrompt} ${customPrompt}` 
+    : defaultPrompt;
+  
+  // 调用阿里生图 API（使用现有的 alibaba-image provider）
+  // 输入：[productImage, sceneImage] + prompt
+  // 输出：融合后的图片
+}
+
+// 批量并发处理所有场景
+async function processAllScenes(
+  productImage: string,
+  scenes: SceneItem[],
+  onComplete: (results: SceneItem[]) => void
+) {
+  // 1. 显示动态占位卡片（使用现有的 ImageGenerationLoading 组件）
+  // 2. 并发发出所有请求
+  const promises = scenes.map(async (scene, index) => {
+    try {
+      const result = await generateSceneFusion(
+        productImage,
+        scene.image,
+        scene.customPrompt  // 使用该场景独立的补充关键词
+      );
+      return { ...scene, status: 'completed' as const, result };
+    } catch (error) {
+      return { ...scene, status: 'failed' as const, error: error.message };
+    }
+  });
+  
+  // 3. 等待所有请求完成
+  const results = await Promise.all(promises);
+  
+  // 4. 同时显示所有结果
+  onComplete(results);
+  
+  // 5. 保存到用户相册和生图记录
+  await saveToGalleryAndRecords(results);
+}
+
+// 保存结果到相册和生图记录
+async function saveToGalleryAndRecords(results: SceneItem[]) {
+  const successResults = results.filter(r => r.status === 'completed' && r.result);
+  
+  for (const result of successResults) {
+    // 调用现有的保存图片 API
+    await fetch('/api/gallery', {
+      method: 'POST',
+      body: JSON.stringify({
+        imageUrl: result.result,
+        // ... 其他参数
+      })
+    });
+  }
+}
+```
+
+#### 4. 批量处理策略
+- **并发处理**：同时发出所有请求，提高处理速度
+- **等待显示**：等待所有图片生成完成后，同时显示结果
+- **加载状态**：使用现有的动态占位卡片显示等待状态
+- **错误处理**：单张失败不影响其他图片，显示失败原因，支持重新生成
+- **结果保存**：✅ 生成的图片保存到用户相册和生图记录
+
+#### 4. 入口位置
+- 图片编辑器页面顶部工具栏添加「场景裂变」按钮
+- 点击打开场景裂变弹窗
+
+---
+
+## 文件修改清单
+
+### 功能一：图片编辑器入口优化
+| 文件 | 修改内容 |
+|------|----------|
+| `components/image-annotator.tsx` | 支持无图片时显示上传界面 |
+| `components/multimodal-input.tsx` | 修改图片编辑器按钮点击逻辑 |
+
+### 功能二：场景裂变
+| 文件 | 修改内容 |
+|------|----------|
+| `components/scene-fusion-dialog.tsx` | 新建场景裂变弹窗组件 |
+| `components/image-annotator.tsx` | 添加「场景裂变」按钮入口 |
+| `app/api/scene-fusion/route.ts` | 新建场景裂变 API（可选，也可直接前端调用现有 chat API） |
+
+---
+
+## 时间估算
+
+| 功能 | 预计工时 |
+|------|----------|
+| 功能一：图片编辑器入口优化 | 1-2 小时 |
+| 功能二：场景裂变 | 3-4 小时 |
+| 测试和调试 | 1 小时 |
+| **总计** | **5-7 小时** |
+
+---
+
+## 待确认事项
+
+1. **场景裂变的默认提示词**：
+   - 当前设计：`将图1的产品替换到图2的场景中，保持图1产品的外观、颜色和细节不变，自然融合到图2的环境中。`
+
+2. **每个场景的补充关键词**：
+   - ✅ 已确认：每个场景图独立配置补充关键词
+   - 用户可以为每个场景单独填写补充描述
+
+3. **最大场景图数量**：
+   - ✅ 已确认：最多 10 张
+
+4. **生成图片数量**：
+   - 当前设计：每个场景生成 1 张
+
+5. **并发处理**：
+   - ✅ 已确认：并发处理（同时发出所有请求）
+   - 等待所有图片生成完成后，同时显示结果
+   - 等待过程中使用现有的动态占位卡片
+
+6. **结果保存**：
+   - ✅ 已确认：保存到用户相册
+   - ✅ 已确认：保存到生图记录
+
+7. **入口位置**：
+   - ✅ 已确认：场景裂变入口放在图片编辑器工具栏中
+
+8. **重新生成**：
+   - ✅ 已确认：支持对单个失败/不满意的场景重新生成
+
+---
+
+请确认以上规划，确认后我将开始开发。