Pointnet_Pointnet2_pytorch

Pointnet_Pointnet2_pytorch copied to clipboard

Why is the Z axis not considered when extracting a random block in the data loader?

while (True):
            center = points[np.random.choice(N_points)][:3]
            block_min = center - [self.block_size / 2.0, self.block_size / 2.0, 0]
            block_max = center + [self.block_size / 2.0, self.block_size / 2.0, 0]
            point_idxs = np.where((points[:, 0] >= block_min[0]) & \
                                  (points[:, 0] <= block_max[0]) & \
                                  (points[:, 1] >= block_min[1]) & \
                                  (points[:, 1] <= block_max[1]))[0]