@lumy-pack/scene-sieve 0.0.5 → 0.0.6

package/dist/cli.mjs

@@ -51,7 +51,7 @@ import { join } from "path";
 function getTempWorkspaceDir(sessionId) {
   return join(TEMP_BASE_DIR, `${WORKSPACE_PREFIX}${sessionId}`);
 }
-var APP_NAME, DEFAULT_COUNT, DEFAULT_THRESHOLD, DEFAULT_FPS, DEFAULT_SCALE, DEFAULT_QUALITY, DEFAULT_MAX_FRAMES, NORMALIZATION_PERCENTILE, WORKSPACE_PREFIX, TEMP_BASE_DIR, SUPPORTED_VIDEO_EXTENSIONS, SUPPORTED_GIF_EXTENSIONS, FRAME_OUTPUT_EXTENSION, OPENCV_BATCH_SIZE, DBSCAN_ALPHA, DBSCAN_MIN_PTS, IOU_THRESHOLD, DECAY_LAMBDA, ANIMATION_FRAME_THRESHOLD, MATCH_DISTANCE_THRESHOLD, PIXELDIFF_GAUSSIAN_KERNEL, PIXELDIFF_BINARY_THRESHOLD, PIXELDIFF_CONTOUR_MIN_AREA, PIXELDIFF_SAMPLE_SPACING;
+var APP_NAME, DEFAULT_COUNT, DEFAULT_THRESHOLD, DEFAULT_FPS, DEFAULT_SCALE, DEFAULT_QUALITY, DEFAULT_MAX_FRAMES, NORMALIZATION_PERCENTILE, WORKSPACE_PREFIX, TEMP_BASE_DIR, SUPPORTED_VIDEO_EXTENSIONS, SUPPORTED_GIF_EXTENSIONS, FRAME_OUTPUT_EXTENSION, FRAME_FILENAME_PATTERN, OPENCV_BATCH_SIZE, DBSCAN_ALPHA, DBSCAN_MIN_PTS, IOU_THRESHOLD, DECAY_LAMBDA, ANIMATION_FRAME_THRESHOLD, MATCH_DISTANCE_THRESHOLD, PIXELDIFF_GAUSSIAN_KERNEL, PIXELDIFF_BINARY_THRESHOLD, PIXELDIFF_CONTOUR_MIN_AREA, PIXELDIFF_SAMPLE_SPACING;
 var init_constants = __esm({
   "src/constants.ts"() {
     "use strict";
@@ -74,6 +74,7 @@ var init_constants = __esm({
     ];
     SUPPORTED_GIF_EXTENSIONS = [".gif"];
     FRAME_OUTPUT_EXTENSION = ".jpg";
+    FRAME_FILENAME_PATTERN = "frame_%06d.jpg";
     OPENCV_BATCH_SIZE = 10;
     DBSCAN_ALPHA = 0.03;
     DBSCAN_MIN_PTS = 4;
@@ -431,13 +432,17 @@ async function analyzeBatch(cvLib, frames, scale, tracker, pairOffset) {
 }
 async function analyzeFrames(ctx) {
   const { frames } = ctx;
-  if (frames.length < 2) return [];
+  if (frames.length < 2) return { edges: [], animations: [] };
   logger.debug(
     `Analyzing ${frames.length} frames in batches of ${OPENCV_BATCH_SIZE}`
   );
   const cvLib = await ensureOpenCV();
   const edges = [];
-  const tracker = new IoUTracker();
+  const tracker = new IoUTracker(
+    ctx.options.fps,
+    ctx.options.iouThreshold,
+    ctx.options.animationThreshold
+  );
   const scale = ctx.options.scale;
   for (let i = 0; i < frames.length - 1; i += OPENCV_BATCH_SIZE) {
     const batchEnd = Math.min(i + OPENCV_BATCH_SIZE + 1, frames.length);
@@ -450,8 +455,11 @@ async function analyzeFrames(ctx) {
     );
     ctx.emitProgress(progress);
   }
-  logger.debug(`Computed ${edges.length} score edges`);
-  return edges;
+  const animations = tracker.flushAndGetAnimations();
+  logger.debug(
+    `Computed ${edges.length} score edges and ${animations.length} animations`
+  );
+  return { edges, animations };
 }
 var OPENCV_INIT_TIMEOUT_MS, require2, cvReady, IoUTracker;
 var init_analyzer = __esm({
@@ -464,7 +472,13 @@ var init_analyzer = __esm({
     require2 = createRequire(import.meta.url);
     cvReady = null;
     IoUTracker = class {
+      constructor(fps = DEFAULT_FPS, iouThreshold = IOU_THRESHOLD, animationThreshold = ANIMATION_FRAME_THRESHOLD) {
+        this.fps = fps;
+        this.iouThreshold = iouThreshold;
+        this.animationThreshold = animationThreshold;
+      }
       regions = [];
+      extractedAnimations = [];
       update(boxes, pairIndex) {
         const animationIndices = /* @__PURE__ */ new Set();
         const matched = /* @__PURE__ */ new Set();
@@ -480,7 +494,7 @@ var init_analyzer = __esm({
               bestRegionIdx = ri;
             }
           }
-          if (bestIoU > IOU_THRESHOLD && bestRegionIdx !== -1) {
+          if (bestIoU > this.iouThreshold && bestRegionIdx !== -1) {
             const region = this.regions[bestRegionIdx];
             const gap = pairIndex - region.lastSeen;
             region.box = box;
@@ -488,13 +502,14 @@ var init_analyzer = __esm({
             region.lastSeen = pairIndex;
             region.weight *= Math.pow(DECAY_LAMBDA, gap);
             matched.add(bestRegionIdx);
-            if (region.consecutiveCount >= ANIMATION_FRAME_THRESHOLD) {
+            if (region.consecutiveCount >= this.animationThreshold) {
               animationIndices.add(bi);
             }
           } else {
             this.regions.push({
               box,
               consecutiveCount: 1,
+              firstSeen: pairIndex,
               lastSeen: pairIndex,
               weight: 1
             });
@@ -506,17 +521,45 @@ var init_analyzer = __esm({
             this.regions[ri].weight *= Math.pow(DECAY_LAMBDA, gap);
           }
         }
-        this.regions = this.regions.filter((r) => r.weight > 0.01);
+        for (let i = 0; i < this.regions.length; i++) {
+          const region = this.regions[i];
+          if (region.weight <= 0.01 && !matched.has(i)) {
+            this.collectAnimation(region);
+          }
+        }
+        this.regions = this.regions.filter(
+          (r, i) => r.weight > 0.01 || matched.has(i)
+        );
         return animationIndices;
       }
+      collectAnimation(region) {
+        if (region.consecutiveCount >= this.animationThreshold) {
+          const durationMs = region.consecutiveCount / this.fps * 1e3;
+          this.extractedAnimations.push({
+            type: "loading_spinner",
+            // 기본값으로 loading_spinner 사용
+            boundingBox: region.box,
+            startFrameId: region.firstSeen,
+            endFrameId: region.lastSeen,
+            durationMs
+          });
+        }
+      }
+      flushAndGetAnimations() {
+        for (const region of this.regions) {
+          this.collectAnimation(region);
+        }
+        this.regions = [];
+        return this.extractedAnimations;
+      }
       getAnimationWeight(boxIndex, boxes) {
         if (boxIndex >= boxes.length) return 0;
         const box = boxes[boxIndex];
         let maxWeight = 0;
         for (const region of this.regions) {
-          if (region.consecutiveCount >= ANIMATION_FRAME_THRESHOLD) {
+          if (region.consecutiveCount >= this.animationThreshold) {
             const iou = computeIoU(box, region.box);
-            if (iou > IOU_THRESHOLD) {
+            if (iou > this.iouThreshold) {
               maxWeight = Math.max(maxWeight, region.weight);
             }
           }
@@ -607,7 +650,7 @@ async function extractFrames(ctx) {
   return frames;
 }
 async function extractByFps(inputPath, outputDir, fps, scale) {
-  const outputPattern = join2(outputDir, "frame_%06d.jpg");
+  const outputPattern = join2(outputDir, FRAME_FILENAME_PATTERN);
   await execa(ffmpegPath, [
     "-i",
     inputPath,
@@ -675,13 +718,44 @@ async function finalizeOutput(ctx, selectedFrames) {
   const outputPath = ctx.options.outputPath;
   const quality = ctx.options.quality;
   const outputFiles = [];
+  const framesMetadata = [];
+  const totalFramesCount = ctx.frames.length;
+  const padding = Math.max(4, String(totalFramesCount).length);
   for (let i = 0; i < selectedFrames.length; i++) {
     const frame = selectedFrames[i];
-    const destName = `scene_${String(i + 1).padStart(3, "0")}.jpg`;
-    const destPath = join3(stagingDir, destName);
+    const fileName = `frame_${String(frame.id + 1).padStart(padding, "0")}.jpg`;
+    const destPath = join3(stagingDir, fileName);
     await sharp2(frame.extractPath).jpeg({ quality, mozjpeg: true }).toFile(destPath);
-    outputFiles.push(join3(outputPath, destName));
+    outputFiles.push(join3(outputPath, fileName));
+    framesMetadata.push({
+      step: i + 1,
+      fileName,
+      frameId: frame.id + 1,
+      timestampMs: Math.round(frame.timestamp * 1e3)
+    });
   }
+  const metadata = {
+    video: {
+      originalDurationMs: Math.round(
+        (ctx.frames.length > 0 ? ctx.frames[ctx.frames.length - 1].timestamp : 0) * 1e3
+      ),
+      fps: ctx.options.fps,
+      resolution: {
+        width: ctx.options.scale,
+        height: Math.round(ctx.options.scale * 9 / 16)
+      }
+    },
+    frames: framesMetadata,
+    animations: (ctx.animations || []).map((anim) => ({
+      ...anim,
+      startFrameId: anim.startFrameId + 1,
+      endFrameId: anim.endFrameId + 1,
+      durationMs: Math.round(anim.durationMs)
+    }))
+  };
+  const metadataPath = join3(stagingDir, ".metadata.json");
+  await writeFile(metadataPath, JSON.stringify(metadata, null, 2));
+  outputFiles.push(join3(outputPath, ".metadata.json"));
   await ensureDir(join3(outputPath, ".."));
   await rm(outputPath, { recursive: true, force: true });
   await rename(stagingDir, outputPath);
@@ -769,6 +843,8 @@ function resolveOptions(options) {
     maxFrames: options.maxFrames ?? DEFAULT_MAX_FRAMES,
     scale: options.scale ?? DEFAULT_SCALE,
     quality: options.quality ?? DEFAULT_QUALITY,
+    iouThreshold: options.iouThreshold ?? IOU_THRESHOLD,
+    animationThreshold: options.animationThreshold ?? ANIMATION_FRAME_THRESHOLD,
     debug: options.debug ?? false
   };
 }
@@ -1065,7 +1141,9 @@ async function runPipeline(options) {
     }
     ctx.emitProgress(100);
     ctx.status = "ANALYZING";
-    ctx.graph = await analyzeFrames(ctx);
+    const { edges, animations } = await analyzeFrames(ctx);
+    ctx.graph = edges;
+    ctx.animations = animations;
     ctx.status = "PRUNING";
     const survivingIds = pruneByThresholdWithCap(
       ctx.graph,
@@ -1098,6 +1176,15 @@ async function runPipeline(options) {
       prunedFramesCount: prunedFrames.length,
       outputFiles,
       outputBuffers,
+      animations: ctx.animations,
+      video: {
+        originalDurationMs: ctx.frames.length / ctx.options.fps * 1e3,
+        fps: ctx.options.fps,
+        resolution: {
+          width: ctx.options.scale,
+          height: Math.round(ctx.options.scale * 9 / 16)
+        }
+      },
       executionTimeMs: Date.now() - startTime
     };
   } catch (error) {
@@ -1280,6 +1367,8 @@ var SieveView = (props) => {
             maxFrames: props.maxFrames,
             scale: props.scale,
             quality: props.quality,
+            iouThreshold: props.iouThreshold,
+            animationThreshold: props.animationThreshold,
             debug: props.debug
           },
           (phase, percent) => {
@@ -1374,12 +1463,19 @@ var SieveView = (props) => {
         "\u2713 Done",
         " \u2014 ",
         result.originalFramesCount,
-        " frames \u2192 ",
+        " frames \u2192",
+        " ",
         result.prunedFramesCount,
         " scenes (",
         (result.executionTimeMs / 1e3).toFixed(1),
         "s)"
       ] }),
+      result.animations && result.animations.length > 0 && /* @__PURE__ */ jsxs3(Text3, { color: "blue", children: [
+        "\u2139 Found",
+        " ",
+        result.animations.length,
+        " animations (recorded in .metadata.json)"
+      ] }),
       props.debug && result.outputFiles.length > 0 && /* @__PURE__ */ jsxs3(Box2, { flexDirection: "column", marginTop: 1, children: [
         /* @__PURE__ */ jsxs3(Text3, { color: "gray", children: [
           "Output: ",
@@ -1395,17 +1491,32 @@ var SieveView = (props) => {
 };
 
 // src/cli.ts
+init_constants();
 var require3 = createRequire2(import.meta.url);
 var { version } = require3("../package.json");
 var program = new Command();
 program.name("scene-sieve").description("Extract key frames from video and GIF files").version(version).argument("<input>", "Input video or GIF file path").option("-n, --count <number>", "Max number of frames to keep (default: 20)").option(
   "-t, --threshold <number>",
   "Normalized threshold 0~1 (default: 0.5; keeps frames above ratio of max change)"
-).option("-o, --output <path>", "Output directory path").option("--fps <number>", "Max FPS for frame extraction", "5").option(
-  "--max-frames <number>",
+).option("-o, --output <path>", "Output directory path").option("--fps <number>", "Max FPS for frame extraction", String(DEFAULT_FPS)).option(
+  "-mf, --max-frames <number>",
   "Max frames to extract (auto-reduces FPS for long videos)",
-  "300"
-).option("-s, --scale <number>", "Scale size for vision analysis", "720").option("-q, --quality <number>", "JPEG output quality 1-100", "80").option("--debug", "Enable debug mode (preserve temp workspace)").action(async (input, opts) => {
+  String(DEFAULT_MAX_FRAMES)
+).option(
+  "-s, --scale <number>",
+  "Scale size for vision analysis",
+  String(DEFAULT_SCALE)
+).option(
+  "-q, --quality <number>",
+  "JPEG output quality 1-100",
+  String(DEFAULT_QUALITY)
+).option(
+  "-it, --iou-threshold <number>",
+  `IoU threshold for animation tracking (0-1) (default: ${IOU_THRESHOLD})`
+).option(
+  "-at, --anim-threshold <number>",
+  `Min consecutive frames for animation (default: ${ANIMATION_FRAME_THRESHOLD})`
+).option("--debug", "Enable debug mode (preserve temp workspace)").action(async (input, opts) => {
   const { waitUntilExit } = render(
     React2.createElement(SieveView, {
       input,
@@ -1416,6 +1527,8 @@ program.name("scene-sieve").description("Extract key frames from video and GIF f
       maxFrames: parseInt(opts.maxFrames, 10),
       scale: parseInt(opts.scale, 10),
       quality: parseInt(opts.quality, 10),
+      iouThreshold: opts.iouThreshold !== void 0 ? parseFloat(opts.iouThreshold) : void 0,
+      animationThreshold: opts.animThreshold !== void 0 ? parseInt(opts.animThreshold, 10) : void 0,
       debug: opts.debug ?? false
     })
   );

package/dist/commands/Sieve.d.ts

@@ -8,6 +8,8 @@ export interface SieveViewProps {
     maxFrames: number;
     scale: number;
     quality: number;
+    iouThreshold?: number;
+    animationThreshold?: number;
     debug: boolean;
 }
 export declare const SieveView: React.FC<SieveViewProps>;

package/dist/core/analyzer.d.ts

@@ -1,4 +1,4 @@
-import type { BoundingBox, ProcessContext, ScoreEdge } from '../types/index.js';
+import type { AnalysisResult, AnimationMetadata, BoundingBox, ProcessContext } from '../types/index.js';
 import type { Point2D } from './dbscan.js';
 type CvLib = typeof import('@techstark/opencv-js');
 export declare function preprocessFrame(framePath: string, scale: number): Promise<{
@@ -8,8 +8,15 @@ export declare function preprocessFrame(framePath: string, scale: number): Promi
 }>;
 export declare function computeIoU(a: BoundingBox, b: BoundingBox): number;
 export declare class IoUTracker {
+    private fps;
+    private iouThreshold;
+    private animationThreshold;
     private regions;
+    private extractedAnimations;
+    constructor(fps?: number, iouThreshold?: number, animationThreshold?: number);
     update(boxes: BoundingBox[], pairIndex: number): Set<number>;
+    private collectAnimation;
+    flushAndGetAnimations(): AnimationMetadata[];
     getAnimationWeight(boxIndex: number, boxes: BoundingBox[]): number;
 }
 export interface AKAZEResult {
@@ -51,5 +58,5 @@ export declare function computeInformationGain(clusters: BoundingBox[], clusterP
  * 3. Spatio-temporal IoU Tracking
  * 4. G(t) Information Gain Scoring
  */
-export declare function analyzeFrames(ctx: ProcessContext): Promise<ScoreEdge[]>;
+export declare function analyzeFrames(ctx: ProcessContext): Promise<AnalysisResult>;
 export {};

package/dist/index.cjs

@@ -97,6 +97,7 @@ var SUPPORTED_VIDEO_EXTENSIONS = [
 ];
 var SUPPORTED_GIF_EXTENSIONS = [".gif"];
 var FRAME_OUTPUT_EXTENSION = ".jpg";
+var FRAME_FILENAME_PATTERN = "frame_%06d.jpg";
 var OPENCV_BATCH_SIZE = 10;
 var DBSCAN_ALPHA = 0.03;
 var DBSCAN_MIN_PTS = 4;
@@ -236,7 +237,13 @@ function computeIoU(a, b) {
   return union === 0 ? 0 : intersection / union;
 }
 var IoUTracker = class {
+  constructor(fps = DEFAULT_FPS, iouThreshold = IOU_THRESHOLD, animationThreshold = ANIMATION_FRAME_THRESHOLD) {
+    this.fps = fps;
+    this.iouThreshold = iouThreshold;
+    this.animationThreshold = animationThreshold;
+  }
   regions = [];
+  extractedAnimations = [];
   update(boxes, pairIndex) {
     const animationIndices = /* @__PURE__ */ new Set();
     const matched = /* @__PURE__ */ new Set();
@@ -252,7 +259,7 @@ var IoUTracker = class {
           bestRegionIdx = ri;
         }
       }
-      if (bestIoU > IOU_THRESHOLD && bestRegionIdx !== -1) {
+      if (bestIoU > this.iouThreshold && bestRegionIdx !== -1) {
         const region = this.regions[bestRegionIdx];
         const gap = pairIndex - region.lastSeen;
         region.box = box;
@@ -260,13 +267,14 @@ var IoUTracker = class {
         region.lastSeen = pairIndex;
         region.weight *= Math.pow(DECAY_LAMBDA, gap);
         matched.add(bestRegionIdx);
-        if (region.consecutiveCount >= ANIMATION_FRAME_THRESHOLD) {
+        if (region.consecutiveCount >= this.animationThreshold) {
           animationIndices.add(bi);
         }
       } else {
         this.regions.push({
           box,
           consecutiveCount: 1,
+          firstSeen: pairIndex,
           lastSeen: pairIndex,
           weight: 1
         });
@@ -278,17 +286,45 @@ var IoUTracker = class {
         this.regions[ri].weight *= Math.pow(DECAY_LAMBDA, gap);
       }
     }
-    this.regions = this.regions.filter((r) => r.weight > 0.01);
+    for (let i = 0; i < this.regions.length; i++) {
+      const region = this.regions[i];
+      if (region.weight <= 0.01 && !matched.has(i)) {
+        this.collectAnimation(region);
+      }
+    }
+    this.regions = this.regions.filter(
+      (r, i) => r.weight > 0.01 || matched.has(i)
+    );
     return animationIndices;
   }
+  collectAnimation(region) {
+    if (region.consecutiveCount >= this.animationThreshold) {
+      const durationMs = region.consecutiveCount / this.fps * 1e3;
+      this.extractedAnimations.push({
+        type: "loading_spinner",
+        // 기본값으로 loading_spinner 사용
+        boundingBox: region.box,
+        startFrameId: region.firstSeen,
+        endFrameId: region.lastSeen,
+        durationMs
+      });
+    }
+  }
+  flushAndGetAnimations() {
+    for (const region of this.regions) {
+      this.collectAnimation(region);
+    }
+    this.regions = [];
+    return this.extractedAnimations;
+  }
   getAnimationWeight(boxIndex, boxes) {
     if (boxIndex >= boxes.length) return 0;
     const box = boxes[boxIndex];
     let maxWeight = 0;
     for (const region of this.regions) {
-      if (region.consecutiveCount >= ANIMATION_FRAME_THRESHOLD) {
+      if (region.consecutiveCount >= this.animationThreshold) {
         const iou = computeIoU(box, region.box);
-        if (iou > IOU_THRESHOLD) {
+        if (iou > this.iouThreshold) {
           maxWeight = Math.max(maxWeight, region.weight);
         }
       }
@@ -510,13 +546,17 @@ async function analyzeBatch(cvLib, frames, scale, tracker, pairOffset) {
 }
 async function analyzeFrames(ctx) {
   const { frames } = ctx;
-  if (frames.length < 2) return [];
+  if (frames.length < 2) return { edges: [], animations: [] };
   logger.debug(
     `Analyzing ${frames.length} frames in batches of ${OPENCV_BATCH_SIZE}`
   );
   const cvLib = await ensureOpenCV();
   const edges = [];
-  const tracker = new IoUTracker();
+  const tracker = new IoUTracker(
+    ctx.options.fps,
+    ctx.options.iouThreshold,
+    ctx.options.animationThreshold
+  );
   const scale = ctx.options.scale;
   for (let i = 0; i < frames.length - 1; i += OPENCV_BATCH_SIZE) {
     const batchEnd = Math.min(i + OPENCV_BATCH_SIZE + 1, frames.length);
@@ -529,8 +569,11 @@ async function analyzeFrames(ctx) {
     );
     ctx.emitProgress(progress);
   }
-  logger.debug(`Computed ${edges.length} score edges`);
-  return edges;
+  const animations = tracker.flushAndGetAnimations();
+  logger.debug(
+    `Computed ${edges.length} score edges and ${animations.length} animations`
+  );
+  return { edges, animations };
 }
 
 // src/core/extractor.ts
@@ -610,7 +653,7 @@ async function extractFrames(ctx) {
   return frames;
 }
 async function extractByFps(inputPath, outputDir, fps, scale) {
-  const outputPattern = (0, import_node_path3.join)(outputDir, "frame_%06d.jpg");
+  const outputPattern = (0, import_node_path3.join)(outputDir, FRAME_FILENAME_PATTERN);
   await (0, import_execa.execa)(import_ffmpeg_static.default, [
     "-i",
     inputPath,
@@ -673,13 +716,44 @@ async function finalizeOutput(ctx, selectedFrames) {
   const outputPath = ctx.options.outputPath;
   const quality = ctx.options.quality;
   const outputFiles = [];
+  const framesMetadata = [];
+  const totalFramesCount = ctx.frames.length;
+  const padding = Math.max(4, String(totalFramesCount).length);
   for (let i = 0; i < selectedFrames.length; i++) {
     const frame = selectedFrames[i];
-    const destName = `scene_${String(i + 1).padStart(3, "0")}.jpg`;
-    const destPath = (0, import_node_path4.join)(stagingDir, destName);
+    const fileName = `frame_${String(frame.id + 1).padStart(padding, "0")}.jpg`;
+    const destPath = (0, import_node_path4.join)(stagingDir, fileName);
     await (0, import_sharp2.default)(frame.extractPath).jpeg({ quality, mozjpeg: true }).toFile(destPath);
-    outputFiles.push((0, import_node_path4.join)(outputPath, destName));
+    outputFiles.push((0, import_node_path4.join)(outputPath, fileName));
+    framesMetadata.push({
+      step: i + 1,
+      fileName,
+      frameId: frame.id + 1,
+      timestampMs: Math.round(frame.timestamp * 1e3)
+    });
   }
+  const metadata = {
+    video: {
+      originalDurationMs: Math.round(
+        (ctx.frames.length > 0 ? ctx.frames[ctx.frames.length - 1].timestamp : 0) * 1e3
+      ),
+      fps: ctx.options.fps,
+      resolution: {
+        width: ctx.options.scale,
+        height: Math.round(ctx.options.scale * 9 / 16)
+      }
+    },
+    frames: framesMetadata,
+    animations: (ctx.animations || []).map((anim) => ({
+      ...anim,
+      startFrameId: anim.startFrameId + 1,
+      endFrameId: anim.endFrameId + 1,
+      durationMs: Math.round(anim.durationMs)
+    }))
+  };
+  const metadataPath = (0, import_node_path4.join)(stagingDir, ".metadata.json");
+  await (0, import_promises3.writeFile)(metadataPath, JSON.stringify(metadata, null, 2));
+  outputFiles.push((0, import_node_path4.join)(outputPath, ".metadata.json"));
   await ensureDir((0, import_node_path4.join)(outputPath, ".."));
   await (0, import_promises3.rm)(outputPath, { recursive: true, force: true });
   await (0, import_promises3.rename)(stagingDir, outputPath);
@@ -743,6 +817,8 @@ function resolveOptions(options) {
     maxFrames: options.maxFrames ?? DEFAULT_MAX_FRAMES,
     scale: options.scale ?? DEFAULT_SCALE,
     quality: options.quality ?? DEFAULT_QUALITY,
+    iouThreshold: options.iouThreshold ?? IOU_THRESHOLD,
+    animationThreshold: options.animationThreshold ?? ANIMATION_FRAME_THRESHOLD,
     debug: options.debug ?? false
   };
 }
@@ -1013,7 +1089,9 @@ async function runPipeline(options) {
     }
     ctx.emitProgress(100);
     ctx.status = "ANALYZING";
-    ctx.graph = await analyzeFrames(ctx);
+    const { edges, animations } = await analyzeFrames(ctx);
+    ctx.graph = edges;
+    ctx.animations = animations;
     ctx.status = "PRUNING";
     const survivingIds = pruneByThresholdWithCap(
       ctx.graph,
@@ -1046,6 +1124,15 @@ async function runPipeline(options) {
       prunedFramesCount: prunedFrames.length,
       outputFiles,
       outputBuffers,
+      animations: ctx.animations,
+      video: {
+        originalDurationMs: ctx.frames.length / ctx.options.fps * 1e3,
+        fps: ctx.options.fps,
+        resolution: {
+          width: ctx.options.scale,
+          height: Math.round(ctx.options.scale * 9 / 16)
+        }
+      },
       executionTimeMs: Date.now() - startTime
     };
   } catch (error) {

package/dist/index.mjs

@@ -57,6 +57,7 @@ var SUPPORTED_VIDEO_EXTENSIONS = [
 ];
 var SUPPORTED_GIF_EXTENSIONS = [".gif"];
 var FRAME_OUTPUT_EXTENSION = ".jpg";
+var FRAME_FILENAME_PATTERN = "frame_%06d.jpg";
 var OPENCV_BATCH_SIZE = 10;
 var DBSCAN_ALPHA = 0.03;
 var DBSCAN_MIN_PTS = 4;
@@ -196,7 +197,13 @@ function computeIoU(a, b) {
   return union === 0 ? 0 : intersection / union;
 }
 var IoUTracker = class {
+  constructor(fps = DEFAULT_FPS, iouThreshold = IOU_THRESHOLD, animationThreshold = ANIMATION_FRAME_THRESHOLD) {
+    this.fps = fps;
+    this.iouThreshold = iouThreshold;
+    this.animationThreshold = animationThreshold;
+  }
   regions = [];
+  extractedAnimations = [];
   update(boxes, pairIndex) {
     const animationIndices = /* @__PURE__ */ new Set();
     const matched = /* @__PURE__ */ new Set();
@@ -212,7 +219,7 @@ var IoUTracker = class {
           bestRegionIdx = ri;
         }
       }
-      if (bestIoU > IOU_THRESHOLD && bestRegionIdx !== -1) {
+      if (bestIoU > this.iouThreshold && bestRegionIdx !== -1) {
         const region = this.regions[bestRegionIdx];
         const gap = pairIndex - region.lastSeen;
         region.box = box;
@@ -220,13 +227,14 @@ var IoUTracker = class {
         region.lastSeen = pairIndex;
         region.weight *= Math.pow(DECAY_LAMBDA, gap);
         matched.add(bestRegionIdx);
-        if (region.consecutiveCount >= ANIMATION_FRAME_THRESHOLD) {
+        if (region.consecutiveCount >= this.animationThreshold) {
           animationIndices.add(bi);
         }
       } else {
         this.regions.push({
           box,
           consecutiveCount: 1,
+          firstSeen: pairIndex,
           lastSeen: pairIndex,
           weight: 1
         });
@@ -238,17 +246,45 @@ var IoUTracker = class {
         this.regions[ri].weight *= Math.pow(DECAY_LAMBDA, gap);
       }
     }
-    this.regions = this.regions.filter((r) => r.weight > 0.01);
+    for (let i = 0; i < this.regions.length; i++) {
+      const region = this.regions[i];
+      if (region.weight <= 0.01 && !matched.has(i)) {
+        this.collectAnimation(region);
+      }
+    }
+    this.regions = this.regions.filter(
+      (r, i) => r.weight > 0.01 || matched.has(i)
+    );
     return animationIndices;
   }
+  collectAnimation(region) {
+    if (region.consecutiveCount >= this.animationThreshold) {
+      const durationMs = region.consecutiveCount / this.fps * 1e3;
+      this.extractedAnimations.push({
+        type: "loading_spinner",
+        // 기본값으로 loading_spinner 사용
+        boundingBox: region.box,
+        startFrameId: region.firstSeen,
+        endFrameId: region.lastSeen,
+        durationMs
+      });
+    }
+  }
+  flushAndGetAnimations() {
+    for (const region of this.regions) {
+      this.collectAnimation(region);
+    }
+    this.regions = [];
+    return this.extractedAnimations;
+  }
   getAnimationWeight(boxIndex, boxes) {
     if (boxIndex >= boxes.length) return 0;
     const box = boxes[boxIndex];
     let maxWeight = 0;
     for (const region of this.regions) {
-      if (region.consecutiveCount >= ANIMATION_FRAME_THRESHOLD) {
+      if (region.consecutiveCount >= this.animationThreshold) {
         const iou = computeIoU(box, region.box);
-        if (iou > IOU_THRESHOLD) {
+        if (iou > this.iouThreshold) {
           maxWeight = Math.max(maxWeight, region.weight);
         }
       }
@@ -470,13 +506,17 @@ async function analyzeBatch(cvLib, frames, scale, tracker, pairOffset) {
 }
 async function analyzeFrames(ctx) {
   const { frames } = ctx;
-  if (frames.length < 2) return [];
+  if (frames.length < 2) return { edges: [], animations: [] };
   logger.debug(
     `Analyzing ${frames.length} frames in batches of ${OPENCV_BATCH_SIZE}`
   );
   const cvLib = await ensureOpenCV();
   const edges = [];
-  const tracker = new IoUTracker();
+  const tracker = new IoUTracker(
+    ctx.options.fps,
+    ctx.options.iouThreshold,
+    ctx.options.animationThreshold
+  );
   const scale = ctx.options.scale;
   for (let i = 0; i < frames.length - 1; i += OPENCV_BATCH_SIZE) {
     const batchEnd = Math.min(i + OPENCV_BATCH_SIZE + 1, frames.length);
@@ -489,8 +529,11 @@ async function analyzeFrames(ctx) {
     );
     ctx.emitProgress(progress);
   }
-  logger.debug(`Computed ${edges.length} score edges`);
-  return edges;
+  const animations = tracker.flushAndGetAnimations();
+  logger.debug(
+    `Computed ${edges.length} score edges and ${animations.length} animations`
+  );
+  return { edges, animations };
 }
 
 // src/core/extractor.ts
@@ -570,7 +613,7 @@ async function extractFrames(ctx) {
   return frames;
 }
 async function extractByFps(inputPath, outputDir, fps, scale) {
-  const outputPattern = join2(outputDir, "frame_%06d.jpg");
+  const outputPattern = join2(outputDir, FRAME_FILENAME_PATTERN);
   await execa(ffmpegPath, [
     "-i",
     inputPath,
@@ -633,13 +676,44 @@ async function finalizeOutput(ctx, selectedFrames) {
   const outputPath = ctx.options.outputPath;
   const quality = ctx.options.quality;
   const outputFiles = [];
+  const framesMetadata = [];
+  const totalFramesCount = ctx.frames.length;
+  const padding = Math.max(4, String(totalFramesCount).length);
   for (let i = 0; i < selectedFrames.length; i++) {
     const frame = selectedFrames[i];
-    const destName = `scene_${String(i + 1).padStart(3, "0")}.jpg`;
-    const destPath = join3(stagingDir, destName);
+    const fileName = `frame_${String(frame.id + 1).padStart(padding, "0")}.jpg`;
+    const destPath = join3(stagingDir, fileName);
     await sharp2(frame.extractPath).jpeg({ quality, mozjpeg: true }).toFile(destPath);
-    outputFiles.push(join3(outputPath, destName));
+    outputFiles.push(join3(outputPath, fileName));
+    framesMetadata.push({
+      step: i + 1,
+      fileName,
+      frameId: frame.id + 1,
+      timestampMs: Math.round(frame.timestamp * 1e3)
+    });
   }
+  const metadata = {
+    video: {
+      originalDurationMs: Math.round(
+        (ctx.frames.length > 0 ? ctx.frames[ctx.frames.length - 1].timestamp : 0) * 1e3
+      ),
+      fps: ctx.options.fps,
+      resolution: {
+        width: ctx.options.scale,
+        height: Math.round(ctx.options.scale * 9 / 16)
+      }
+    },
+    frames: framesMetadata,
+    animations: (ctx.animations || []).map((anim) => ({
+      ...anim,
+      startFrameId: anim.startFrameId + 1,
+      endFrameId: anim.endFrameId + 1,
+      durationMs: Math.round(anim.durationMs)
+    }))
+  };
+  const metadataPath = join3(stagingDir, ".metadata.json");
+  await writeFile(metadataPath, JSON.stringify(metadata, null, 2));
+  outputFiles.push(join3(outputPath, ".metadata.json"));
   await ensureDir(join3(outputPath, ".."));
   await rm(outputPath, { recursive: true, force: true });
   await rename(stagingDir, outputPath);
@@ -703,6 +777,8 @@ function resolveOptions(options) {
     maxFrames: options.maxFrames ?? DEFAULT_MAX_FRAMES,
     scale: options.scale ?? DEFAULT_SCALE,
     quality: options.quality ?? DEFAULT_QUALITY,
+    iouThreshold: options.iouThreshold ?? IOU_THRESHOLD,
+    animationThreshold: options.animationThreshold ?? ANIMATION_FRAME_THRESHOLD,
     debug: options.debug ?? false
   };
 }
@@ -973,7 +1049,9 @@ async function runPipeline(options) {
     }
     ctx.emitProgress(100);
     ctx.status = "ANALYZING";
-    ctx.graph = await analyzeFrames(ctx);
+    const { edges, animations } = await analyzeFrames(ctx);
+    ctx.graph = edges;
+    ctx.animations = animations;
     ctx.status = "PRUNING";
     const survivingIds = pruneByThresholdWithCap(
       ctx.graph,
@@ -1006,6 +1084,15 @@ async function runPipeline(options) {
       prunedFramesCount: prunedFrames.length,
       outputFiles,
       outputBuffers,
+      animations: ctx.animations,
+      video: {
+        originalDurationMs: ctx.frames.length / ctx.options.fps * 1e3,
+        fps: ctx.options.fps,
+        resolution: {
+          width: ctx.options.scale,
+          height: Math.round(ctx.options.scale * 9 / 16)
+        }
+      },
       executionTimeMs: Date.now() - startTime
     };
   } catch (error) {

package/dist/pipeline-worker.mjs

@@ -60,6 +60,7 @@ var SUPPORTED_VIDEO_EXTENSIONS = [
 ];
 var SUPPORTED_GIF_EXTENSIONS = [".gif"];
 var FRAME_OUTPUT_EXTENSION = ".jpg";
+var FRAME_FILENAME_PATTERN = "frame_%06d.jpg";
 var OPENCV_BATCH_SIZE = 10;
 var DBSCAN_ALPHA = 0.03;
 var DBSCAN_MIN_PTS = 4;
@@ -199,7 +200,13 @@ function computeIoU(a, b) {
   return union === 0 ? 0 : intersection / union;
 }
 var IoUTracker = class {
+  constructor(fps = DEFAULT_FPS, iouThreshold = IOU_THRESHOLD, animationThreshold = ANIMATION_FRAME_THRESHOLD) {
+    this.fps = fps;
+    this.iouThreshold = iouThreshold;
+    this.animationThreshold = animationThreshold;
+  }
   regions = [];
+  extractedAnimations = [];
   update(boxes, pairIndex) {
     const animationIndices = /* @__PURE__ */ new Set();
     const matched = /* @__PURE__ */ new Set();
@@ -215,7 +222,7 @@ var IoUTracker = class {
           bestRegionIdx = ri;
         }
       }
-      if (bestIoU > IOU_THRESHOLD && bestRegionIdx !== -1) {
+      if (bestIoU > this.iouThreshold && bestRegionIdx !== -1) {
         const region = this.regions[bestRegionIdx];
         const gap = pairIndex - region.lastSeen;
         region.box = box;
@@ -223,13 +230,14 @@ var IoUTracker = class {
         region.lastSeen = pairIndex;
         region.weight *= Math.pow(DECAY_LAMBDA, gap);
         matched.add(bestRegionIdx);
-        if (region.consecutiveCount >= ANIMATION_FRAME_THRESHOLD) {
+        if (region.consecutiveCount >= this.animationThreshold) {
           animationIndices.add(bi);
         }
       } else {
         this.regions.push({
           box,
           consecutiveCount: 1,
+          firstSeen: pairIndex,
           lastSeen: pairIndex,
           weight: 1
         });
@@ -241,17 +249,45 @@ var IoUTracker = class {
         this.regions[ri].weight *= Math.pow(DECAY_LAMBDA, gap);
       }
     }
-    this.regions = this.regions.filter((r) => r.weight > 0.01);
+    for (let i = 0; i < this.regions.length; i++) {
+      const region = this.regions[i];
+      if (region.weight <= 0.01 && !matched.has(i)) {
+        this.collectAnimation(region);
+      }
+    }
+    this.regions = this.regions.filter(
+      (r, i) => r.weight > 0.01 || matched.has(i)
+    );
     return animationIndices;
   }
+  collectAnimation(region) {
+    if (region.consecutiveCount >= this.animationThreshold) {
+      const durationMs = region.consecutiveCount / this.fps * 1e3;
+      this.extractedAnimations.push({
+        type: "loading_spinner",
+        // 기본값으로 loading_spinner 사용
+        boundingBox: region.box,
+        startFrameId: region.firstSeen,
+        endFrameId: region.lastSeen,
+        durationMs
+      });
+    }
+  }
+  flushAndGetAnimations() {
+    for (const region of this.regions) {
+      this.collectAnimation(region);
+    }
+    this.regions = [];
+    return this.extractedAnimations;
+  }
   getAnimationWeight(boxIndex, boxes) {
     if (boxIndex >= boxes.length) return 0;
     const box = boxes[boxIndex];
     let maxWeight = 0;
     for (const region of this.regions) {
-      if (region.consecutiveCount >= ANIMATION_FRAME_THRESHOLD) {
+      if (region.consecutiveCount >= this.animationThreshold) {
         const iou = computeIoU(box, region.box);
-        if (iou > IOU_THRESHOLD) {
+        if (iou > this.iouThreshold) {
           maxWeight = Math.max(maxWeight, region.weight);
         }
       }
@@ -473,13 +509,17 @@ async function analyzeBatch(cvLib, frames, scale, tracker, pairOffset) {
 }
 async function analyzeFrames(ctx) {
   const { frames } = ctx;
-  if (frames.length < 2) return [];
+  if (frames.length < 2) return { edges: [], animations: [] };
   logger.debug(
     `Analyzing ${frames.length} frames in batches of ${OPENCV_BATCH_SIZE}`
   );
   const cvLib = await ensureOpenCV();
   const edges = [];
-  const tracker = new IoUTracker();
+  const tracker = new IoUTracker(
+    ctx.options.fps,
+    ctx.options.iouThreshold,
+    ctx.options.animationThreshold
+  );
   const scale = ctx.options.scale;
   for (let i = 0; i < frames.length - 1; i += OPENCV_BATCH_SIZE) {
     const batchEnd = Math.min(i + OPENCV_BATCH_SIZE + 1, frames.length);
@@ -492,8 +532,11 @@ async function analyzeFrames(ctx) {
     );
     ctx.emitProgress(progress);
   }
-  logger.debug(`Computed ${edges.length} score edges`);
-  return edges;
+  const animations = tracker.flushAndGetAnimations();
+  logger.debug(
+    `Computed ${edges.length} score edges and ${animations.length} animations`
+  );
+  return { edges, animations };
 }
 
 // src/core/extractor.ts
@@ -573,7 +616,7 @@ async function extractFrames(ctx) {
   return frames;
 }
 async function extractByFps(inputPath, outputDir, fps, scale) {
-  const outputPattern = join2(outputDir, "frame_%06d.jpg");
+  const outputPattern = join2(outputDir, FRAME_FILENAME_PATTERN);
   await execa(ffmpegPath, [
     "-i",
     inputPath,
@@ -636,13 +679,44 @@ async function finalizeOutput(ctx, selectedFrames) {
   const outputPath = ctx.options.outputPath;
   const quality = ctx.options.quality;
   const outputFiles = [];
+  const framesMetadata = [];
+  const totalFramesCount = ctx.frames.length;
+  const padding = Math.max(4, String(totalFramesCount).length);
   for (let i = 0; i < selectedFrames.length; i++) {
     const frame = selectedFrames[i];
-    const destName = `scene_${String(i + 1).padStart(3, "0")}.jpg`;
-    const destPath = join3(stagingDir, destName);
+    const fileName = `frame_${String(frame.id + 1).padStart(padding, "0")}.jpg`;
+    const destPath = join3(stagingDir, fileName);
     await sharp2(frame.extractPath).jpeg({ quality, mozjpeg: true }).toFile(destPath);
-    outputFiles.push(join3(outputPath, destName));
+    outputFiles.push(join3(outputPath, fileName));
+    framesMetadata.push({
+      step: i + 1,
+      fileName,
+      frameId: frame.id + 1,
+      timestampMs: Math.round(frame.timestamp * 1e3)
+    });
   }
+  const metadata = {
+    video: {
+      originalDurationMs: Math.round(
+        (ctx.frames.length > 0 ? ctx.frames[ctx.frames.length - 1].timestamp : 0) * 1e3
+      ),
+      fps: ctx.options.fps,
+      resolution: {
+        width: ctx.options.scale,
+        height: Math.round(ctx.options.scale * 9 / 16)
+      }
+    },
+    frames: framesMetadata,
+    animations: (ctx.animations || []).map((anim) => ({
+      ...anim,
+      startFrameId: anim.startFrameId + 1,
+      endFrameId: anim.endFrameId + 1,
+      durationMs: Math.round(anim.durationMs)
+    }))
+  };
+  const metadataPath = join3(stagingDir, ".metadata.json");
+  await writeFile(metadataPath, JSON.stringify(metadata, null, 2));
+  outputFiles.push(join3(outputPath, ".metadata.json"));
   await ensureDir(join3(outputPath, ".."));
   await rm(outputPath, { recursive: true, force: true });
   await rename(stagingDir, outputPath);
@@ -706,6 +780,8 @@ function resolveOptions(options2) {
     maxFrames: options2.maxFrames ?? DEFAULT_MAX_FRAMES,
     scale: options2.scale ?? DEFAULT_SCALE,
     quality: options2.quality ?? DEFAULT_QUALITY,
+    iouThreshold: options2.iouThreshold ?? IOU_THRESHOLD,
+    animationThreshold: options2.animationThreshold ?? ANIMATION_FRAME_THRESHOLD,
     debug: options2.debug ?? false
   };
 }
@@ -976,7 +1052,9 @@ async function runPipeline(options2) {
     }
     ctx.emitProgress(100);
     ctx.status = "ANALYZING";
-    ctx.graph = await analyzeFrames(ctx);
+    const { edges, animations } = await analyzeFrames(ctx);
+    ctx.graph = edges;
+    ctx.animations = animations;
     ctx.status = "PRUNING";
     const survivingIds = pruneByThresholdWithCap(
       ctx.graph,
@@ -1009,6 +1087,15 @@ async function runPipeline(options2) {
       prunedFramesCount: prunedFrames.length,
       outputFiles,
       outputBuffers,
+      animations: ctx.animations,
+      video: {
+        originalDurationMs: ctx.frames.length / ctx.options.fps * 1e3,
+        fps: ctx.options.fps,
+        resolution: {
+          width: ctx.options.scale,
+          height: Math.round(ctx.options.scale * 9 / 16)
+        }
+      },
       executionTimeMs: Date.now() - startTime
     };
   } catch (error) {

package/dist/types/index.d.ts

@@ -17,6 +17,8 @@ export interface SieveOptionsBase {
     maxFrames?: number;
     scale?: number;
     quality?: number;
+    iouThreshold?: number;
+    animationThreshold?: number;
     debug?: boolean;
     onProgress?: (phase: ProgressPhase, percent: number) => void;
 }
@@ -32,6 +34,8 @@ export interface ResolvedOptions {
     maxFrames: number;
     scale: number;
     quality: number;
+    iouThreshold: number;
+    animationThreshold: number;
     debug: boolean;
 }
 export interface SieveResult {
@@ -40,8 +44,25 @@ export interface SieveResult {
     prunedFramesCount: number;
     outputFiles: string[];
     outputBuffers?: Buffer[];
+    animations?: AnimationMetadata[];
+    video?: VideoMetadata;
     executionTimeMs: number;
 }
+export interface AnimationMetadata {
+    type: string;
+    boundingBox: BoundingBox;
+    startFrameId: number;
+    endFrameId: number;
+    durationMs: number;
+}
+export interface VideoMetadata {
+    originalDurationMs: number;
+    fps: number;
+    resolution: {
+        width: number;
+        height: number;
+    };
+}
 export interface FrameNode {
     id: number;
     timestamp: number;
@@ -75,7 +96,12 @@ export interface ProcessContext {
     workspacePath: string;
     frames: FrameNode[];
     graph: ScoreEdge[];
+    animations?: AnimationMetadata[];
     status: 'INIT' | ProgressPhase | 'SUCCESS' | 'FAILED';
     emitProgress: (percent: number) => void;
     error?: Error;
 }
+export interface AnalysisResult {
+    edges: ScoreEdge[];
+    animations: AnimationMetadata[];
+}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@lumy-pack/scene-sieve",
-  "version": "0.0.5",
+  "version": "0.0.6",
   "description": "CLI tool for extracting key frames from video and GIF files",
   "keywords": [
     "cli",