Convert to Point Cloud

The algorithm converts a depth image into a point cloud

Input

Takes one or two 2D image set(s) as input. The first image set must be representing depth data, i.e. it must contain float single channel images. The second image set is an optional set of corresponding RGB/greyscale images, only data of unsigned char or unsigned short type is supported. Both image sets need a camera calibration data component. When depth and color are provided together, an optional RGB-D calibration data component on the color set supplies the depth-to-color transform; otherwise it is derived from the camera poses.

Output

Point cloud, possibly with colored vertices

Description

The algorithm takes pairs of depth and optionally color images and converts them to point clouds. The depthCutoff parameter specifies the maximum depth value to be considered. Pixels with depth larger than this value are ignored.