Dominant-Flow Visualizer Adjustments

app.py

@@ -13,7 +13,7 @@ def extract_motion_vectors(data):
             continue
         diffs = np.diff(pts, axis=0)
         for d in diffs:
-            if np.linalg.norm(d) > 1:  # ignore jitter
+            if np.linalg.norm(d) > 1:   # ignore jitter / static points
                 vectors.append(d)
     return np.array(vectors)
 
@@ -34,7 +34,8 @@ def learn_flows(vectors, n_clusters=2):
 # 🎨 3. Visualization utility
 # ------------------------------------------------------------
 def draw_flow_overlay(vectors, labels, centers, bg_img=None):
-    if bg_img:
+    # background setup
+    if bg_img and os.path.exists(bg_img):
         bg = cv2.imread(bg_img)
         if bg is None:
             bg = np.ones((600,900,3),dtype=np.uint8)*40
@@ -42,41 +43,60 @@ def draw_flow_overlay(vectors, labels, centers, bg_img=None):
         bg = np.ones((600,900,3),dtype=np.uint8)*40
 
     overlay = bg.copy()
-    colors = [(255,0,0),(0,255,255)]
+    colors = [(0,0,255),(255,255,0)]   # red & yellow for lanes
 
-    for (vx,vy), lab in zip(vectors, labels):
-        start = (np.random.randint(0, overlay.shape[1]), np.random.randint(0, overlay.shape[0]))
-        end = (int(start[0]+vx*2), int(start[1]+vy*2))
+    # Normalize arrow lengths
+    norms = np.linalg.norm(vectors, axis=1, keepdims=True)
+    vectors = np.divide(vectors, norms + 1e-6) * 10
+
+    # Draw sampled small flow lines
+    for i, ((vx,vy), lab) in enumerate(zip(vectors, labels)):
+        if i % 15 != 0:   # draw 1 of every 15 for clarity
+            continue
+        start = (np.random.randint(0, overlay.shape[1]),
+                 np.random.randint(0, overlay.shape[0]))
+        end = (int(start[0]+vx), int(start[1]+vy))
         cv2.arrowedLine(overlay, start, end, colors[lab%2], 1, tipLength=0.3)
 
-    # cluster centroids
+    # Draw two main flow centroids (green)
+    h,w = overlay.shape[:2]
+    center_pt = (w//2, h//2)
     for i, c in enumerate(centers):
-        h,w = overlay.shape[:2]
-        center_pt = (w//2, h//2)
-        tip = (int(center_pt[0]+c[0]*8), int(center_pt[1]+c[1]*8))
-        cv2.arrowedLine(overlay, center_pt, tip, (0,255,0), 3, tipLength=0.4)
-        cv2.putText(overlay, f"Flow {i+1}", (tip[0]+10, tip[1]), cv2.FONT_HERSHEY_SIMPLEX,0.6,(0,255,0),2)
+        end = (int(center_pt[0]+c[0]*40), int(center_pt[1]+c[1]*40))
+        cv2.arrowedLine(overlay, center_pt, end, (0,255,0), 4, tipLength=0.4)
+        cv2.putText(overlay, f"Flow {i+1}", (end[0]+10, end[1]),
+                    cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0,255,0), 2)
+
+    # Blend overlay with background (transparency)
+    alpha = 0.6
+    combined = cv2.addWeighted(bg, alpha, overlay, 1-alpha, 0)
 
-    tmp_path = tempfile.NamedTemporaryFile(suffix=".jpg", delete=False).name
-    cv2.imwrite(tmp_path, overlay)
-    return tmp_path
+    out_path = tempfile.NamedTemporaryFile(suffix=".jpg", delete=False).name
+    cv2.imwrite(out_path, combined)
+    return out_path
 
 
 # ------------------------------------------------------------
 # 🚀 4. Combined pipeline
 # ------------------------------------------------------------
 def process_json(json_file, background=None):
-    data = json.load(open(json_file))
+    try:
+        data = json.load(open(json_file))
+    except Exception as e:
+        return None, {"error": f"Invalid JSON file: {e}"}
+
     vectors = extract_motion_vectors(data)
     if len(vectors)==0:
         return None, {"error":"No motion vectors found."}
+
     labels, centers = learn_flows(vectors)
     if labels is None:
         return None, {"error":"Insufficient data for clustering."}
+
     img_path = draw_flow_overlay(vectors, labels, centers, background)
     stats = {
-        "num_vectors": len(vectors),
-        "dominant_flows": len(centers),
+        "num_vectors": int(len(vectors)),
+        "dominant_flows": int(len(centers)),
         "flow_centers": centers.tolist()
     }
     return img_path, stats
@@ -88,7 +108,7 @@ def process_json(json_file, background=None):
 description_text = """
 ### 🧭 Dominant Flow Learning (Stage 2)
 Upload the **trajectories JSON** from Stage 1 (vehicle tracking).  
-Optionally upload a background road frame image to visualize motion arrows.
+Optionally upload a background road frame image for overlay visualization.
 """
 
 example_json = "trajectories_sample.json" if os.path.exists("trajectories_sample.json") else None