How to create color point cloud from disparity

[Toaster0617]

Required Info
Camera Model	{D400}
Firmware Version	(Open RealSense Viewer --> Click info)
Operating System & Version	Win10
Platform	PC
SDK Version	{ legacy / 2... }
Language	{python/opencv}
Segment	{Robot/VR}

Hi. I'm quite new to using pyrealsense2 library. I having trouble in convert disparity to color point cloud using opencv function (eg. reprojectImageTo3D function).

I've already get my disparity value by using method in #8572 but my disparity value lies between 0 to 65535. So did I need to scale or convert my result so that it can be used for generate color point cloud by using opencv?

Here is the image I get

[MartyG-RealSense]

Hi @Toaster0617 You could convert the RealSense 16-bit uint16_t 'pixel depth value' (which ranges from 0 to 65535) to OpenCV's range of 0-255 by converting the depth value to the OpenCV format CV_16UC1. The #8572 case that you quoted has an example of such a conversion in one of its scripts, at #8572 (comment)

disparity = data.view(np.uint16).reshape((imageH, imageW))

If disparity values are not vital to your project then an alternative approach for obtaining a colorized point cloud with OpenCV code could be the librealsense SDK's opencv_pointcloud_viewer.py pyrealsense2 example program.

https://github.com/IntelRealSense/librealsense/blob/master/wrappers/python/examples/opencv_pointcloud_viewer.py

[Toaster0617]

@MartyG-RealSense Thanks for your help! Does this mean that I only need to convert my data format to 16-bit and scale it to 0-255?

[MartyG-RealSense]

My understanding is that after RealSense data is converted from 16-bit uint16_t to an OpenCV format such as CV_16UC1 then its scale is automatically adjusted without you having to rescale the depth values yourself.

[Toaster0617]

@MartyG-RealSense Thanks for your response! I can solve my problem now!

[MartyG-RealSense]

Case closed due to solution achieved and no further comments received.