Upload 3 files

utils/depth_utils.py

@@ -0,0 +1,25 @@
+import torch
+import cv2
+import numpy as np
+from transformers import DPTFeatureExtractor, DPTForDepthEstimation
+
+feature_extractor = DPTFeatureExtractor.from_pretrained("Intel/dpt-large")
+model = DPTForDepthEstimation.from_pretrained("Intel/dpt-large")
+model.eval()
+
+def estimate_depth(frame_paths):
+    depth_maps = []
+
+    for path in frame_paths:
+        image = cv2.imread(path)
+        image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
+
+        inputs = feature_extractor(images=image, return_tensors="pt")
+        with torch.no_grad():
+            outputs = model(**inputs)
+            depth = outputs.predicted_depth[0].cpu().numpy()
+
+        depth = cv2.resize(depth, (image.shape[1], image.shape[0]))
+        depth_maps.append(depth)
+
+    return depth_maps

utils/pointcloud_utils.py

@@ -0,0 +1,25 @@
+import open3d as o3d
+import numpy as np
+import cv2
+import os
+
+def depth_to_pointcloud(depth_maps, frame_paths, out_dir):
+    points = []
+
+    for depth, frame_path in zip(depth_maps, frame_paths):
+        img = cv2.imread(frame_path)
+        h, w = depth.shape
+
+        for y in range(0, h, 10):
+            for x in range(0, w, 10):
+                z = depth[y, x]
+                points.append([x, y, z])
+
+    points = np.array(points)
+    pcd = o3d.geometry.PointCloud()
+    pcd.points = o3d.utility.Vector3dVector(points)
+
+    out_path = os.path.join(out_dir, "model.ply")
+    o3d.io.write_point_cloud(out_path, pcd)
+
+    return out_path

utils/video_utils.py

@@ -0,0 +1,24 @@
+import cv2
+import os
+
+def extract_frames(video_path, out_dir, every_n=15):
+    cap = cv2.VideoCapture(video_path)
+    frames = []
+    count = 0
+    idx = 0
+
+    while cap.isOpened():
+        ret, frame = cap.read()
+        if not ret:
+            break
+
+        if count % every_n == 0:
+            frame_path = os.path.join(out_dir, f"frame_{idx}.jpg")
+            cv2.imwrite(frame_path, frame)
+            frames.append(frame_path)
+            idx += 1
+
+        count += 1
+
+    cap.release()
+    return frames