ni*SrSSKrSSKrSrSrSrg)z/Common functionalities to process point clouds.NcUS[R"US- 5-- n[R"[R"U5US/5nSUS--U- U- U-S- n[R"[R"U5US/5R nSUS--U- U- U-S- n[R "XhU4SS9n U $)aConverts from a depth map to a point cloud. Args: distances: An numpy array which has the shape of (height, width) that denotes a distance map. The unit is meter. fov: The field of view of the camera in the vertical direction. The unit is radian. width: The width of the image resolution of the camera. height: The height of the image resolution of the camera. Returns: point_cloud: The converted point cloud from the distance map. It is a numpy array of shape (height, width, 3). g@g?axis)mathtannptilearangeTstack) distancesfovwidthheightfpxxpyy point_clouds i/home/james-whalen/.local/lib/python3.13/site-packages/pybullet_envs/minitaur/vision/point_cloud_utils.pydistance_map_to_point_cloudrs DHHS3Y''(! wwryy&!-"BH~"Y.2! wwryy 5!*-//"BH~!#i/!3!! *4+ c[R"USS9n[R"USUS5n[RR USS9SS2SS2S4nX#S-- nU$)zEstimates the normals of a point cloud using cross product. Args: points: a point cloud represented as an 2D array of which has the shape of [m, n, 3]. Returns: normals: a 2D array of normal vectors which has the same shape as the input. )rrrrrrNgHz>)r gradientcrosslinalgnorm)pointsgradnormalnorms normalizeds rpoint_cloud_normalsr' s_ V& )$ 88DGT!W %& ))..a. (At 4%&* rcB[RRX- SS9$)a Estimates the distance of a point to every point in a point cloud. Args: point: a 3D vector representing a point. point_cloud: a point cloud represented as an 2D array of which has the shape of [m, n, 3]. Returns: distances: a distance field represented as a 2D array [m, n]. rr)r r r!)pointrs rpoint_cloud_distancesr*1s  +! 44r)__doc__r numpyr rr'r*rrr.s5 0" 5r