update normal in computing strategy

configs/local/strategy_generate_config.yaml

@@ -13,7 +13,7 @@ runner:
   generate:
     voxel_threshold: 0.003
     overlap_area_threshold: 25
-    compute_with_normal: True
+    compute_with_normal: False
     scan_points_threshold: 10
     overwrite: False
     seq_num: 15

preprocess/preprocessor.py

@@ -164,9 +164,9 @@ def save_scene_data(root, scene, scene_idx=0, scene_total=1,file_type="txt"):
 
 if __name__ == "__main__":
     #root = "/media/hofee/repository/new_data_with_normal"
-    root = r"C:\Document\Local Project\nbv_rec\nbv_reconstruction\temp"
+    root = r"C:\Document\Datasets\nbv_rec_part2"
     scene_list = os.listdir(root)
-    from_idx = 0 # 1000
+    from_idx = 600 # 1000
     to_idx = len(scene_list) # 1500
 
 

utils/pts.py

@@ -7,8 +7,15 @@ class PtsUtil:
     @staticmethod
     def voxel_downsample_point_cloud(point_cloud, voxel_size=0.005, require_idx=False):
         voxel_indices = np.floor(point_cloud / voxel_size).astype(np.int32)
-        unique_voxels = np.unique(voxel_indices, axis=0, return_inverse=True)
+        if require_idx:
-        return unique_voxels[0]*voxel_size
+            _, inverse, counts = np.unique(voxel_indices, axis=0, return_inverse=True, return_counts=True)
+            idx_sort = np.argsort(inverse)
+            idx_unique = idx_sort[np.cumsum(counts)-counts]
+            downsampled_points = point_cloud[idx_unique]
+            return downsampled_points, idx_unique
+        else:
+            unique_voxels = np.unique(voxel_indices, axis=0, return_inverse=True)
+            return unique_voxels[0]*voxel_size
     
     @staticmethod
     def random_downsample_point_cloud(point_cloud, num_points, require_idx=False):

utils/reconstruction.py

@@ -19,6 +19,12 @@ class ReconstructionUtil:
         is_covered_by_distance = distances < threshold*2
         normal_dots = np.einsum('ij,ij->i', target_normal, combined_normal[indices])
         is_covered_by_normal = normal_dots > normal_threshold
+
+        pts_nrm_target = np.hstack([target_point_cloud, target_normal])
+        np.savetxt("pts_nrm_target.txt",  pts_nrm_target)
+        pts_nrm_combined = np.hstack([combined_point_cloud, combined_normal])
+        np.savetxt("pts_nrm_combined.txt", pts_nrm_combined)
+        import ipdb; ipdb.set_trace()
         covered_points_num = np.sum(is_covered_by_distance & is_covered_by_normal)
         coverage_rate = covered_points_num / target_point_cloud.shape[0]
 
@@ -145,7 +151,6 @@ class ReconstructionUtil:
         max_rec_nrm = np.vstack(normal_list)
         downsampled_max_rec_pts, idx = PtsUtil.voxel_downsample_point_cloud(max_rec_pts, threshold, require_idx=True)
         downsampled_max_rec_nrm = max_rec_nrm[idx]  
-        
         max_rec_pts_num = downsampled_max_rec_pts.shape[0]
         try:
             max_real_rec_pts_coverage, _ = ReconstructionUtil.compute_coverage_rate_with_normal(target_point_cloud, downsampled_max_rec_pts, target_normal, downsampled_max_rec_nrm, threshold)
@@ -184,7 +189,7 @@ class ReconstructionUtil:
                 
                 new_combined_point_cloud = np.vstack([combined_point_cloud, point_cloud_list[view_index]])
                 new_combined_normal = np.vstack([combined_normal, normal_list[view_index]])
-                new_downsampled_combined_point_cloud = PtsUtil.voxel_downsample_point_cloud(new_combined_point_cloud,threshold)
+                new_downsampled_combined_point_cloud, idx = PtsUtil.voxel_downsample_point_cloud(new_combined_point_cloud,threshold, require_idx=True)
                 new_downsampled_combined_normal = new_combined_normal[idx]
                 new_coverage, new_covered_num = ReconstructionUtil.compute_coverage_rate_with_normal(downsampled_max_rec_pts, new_downsampled_combined_point_cloud, downsampled_max_rec_nrm, new_downsampled_combined_normal, threshold)
                 coverage_increase = new_coverage - current_coverage

utils/vis.py

@@ -157,6 +157,17 @@ class visualizeUtil:
         np.savetxt(os.path.join(output_dir, "target_pts.txt"), sampled_target_points)
         np.savetxt(os.path.join(output_dir, "target_normal.txt"), sampled_visualized_normal)
 
+    @staticmethod
+    def save_pts_nrm(pts_nrm, output_dir):
+        pts = pts_nrm[:, :3]
+        nrm = pts_nrm[:, 3:]
+        visualized_nrm = []
+        num_samples = 10
+        for i in range(len(pts)):
+            visualized_nrm.append(pts[i] + 0.02*t * nrm[i] for t in range(num_samples))
+        visualized_nrm = np.array(visualized_nrm).reshape(-1, 3)
+        np.savetxt(os.path.join(output_dir, "nrm.txt"), visualized_nrm)
+        np.savetxt(os.path.join(output_dir, "pts.txt"), pts)
 
 
 # ------ Debug ------
@@ -168,8 +179,11 @@ if __name__ == "__main__":
     output_dir = r"C:\Document\Local Project\nbv_rec\nbv_reconstruction\test"
     
     #visualizeUtil.save_all_cam_pos_and_cam_axis(root, scene, output_dir)
-    visualizeUtil.save_all_combined_pts(root, scene, output_dir)
+    # visualizeUtil.save_all_combined_pts(root, scene, output_dir)
-    visualizeUtil.save_seq_combined_pts(root, scene, [0, 121, 286, 175, 111,366,45,230,232,225,255,17,199,78,60], output_dir)
+    # visualizeUtil.save_seq_combined_pts(root, scene, [0, 121, 286, 175, 111,366,45,230,232,225,255,17,199,78,60], output_dir)
-    visualizeUtil.save_seq_cam_pos_and_cam_axis(root, scene, [0, 121, 286, 175, 111,366,45,230,232,225,255,17,199,78,60], output_dir)
+    # visualizeUtil.save_seq_cam_pos_and_cam_axis(root, scene, [0, 121, 286, 175, 111,366,45,230,232,225,255,17,199,78,60], output_dir)
-    visualizeUtil.save_target_mesh_at_world_space(root, model_dir, scene)
+    # visualizeUtil.save_target_mesh_at_world_space(root, model_dir, scene)
     #visualizeUtil.save_points_and_normals(root, scene,"10", output_dir, binocular=True)
+    pts_nrm = np.loadtxt(r"C:\Document\Local Project\nbv_rec\nbv_reconstruction\pts_nrm_target.txt")
+    visualizeUtil.save_pts_nrm(pts_nrm, output_dir)
+