Cannot obtain RGB video stream in YOLOv4 model using RealSense (D435) Camera

[VAIBHAVPATEL97]

---

Required Info
Camera Model	D435
Firmware Version	05.12.14.50
Operating System & Version	Ubuntu 18.04
Kernel Version (Linux Only)	4.9.201-tegra
Platform	NVIDIA Jetson AGX Xavier
SDK Version	v2.48.0
Language	Python
Segment	Others
Jetpack	Jetpack v4.5.1

Hello, I trying to use Intel RealSense D435 camera as an input to the YOLOv4 darknet model but it shows a grayscale version of the captured video. How can I convert the input video stream to an RGB version as a normal camera ?. (Attached screenshot of the input video captured using RealSense).

[MartyG-RealSense]

Hi @VAIBHAVPATEL97 Because of the dot pattern projection on your kindly provided image, it looks as though you have captured the monochrome infrared stream of the camera rather than the camera's RGB stream.

Could you provide information please about how you capture the image (for example, with a Python script). If it is a script, please share the code if possible in a comment below. Thanks!

[VAIBHAVPATEL97]

Hi @MartyG-RealSense, I was able to fix the issue by changing some code in the python script. Another question is it possible to change the resolution of the Intel RealSense during RGB video capturing ?

[MartyG-RealSense]

Hi @VAIBHAVPATEL97 I'm pleased to hear that you resolved the problem with the stream type.

To change the resolution whilst the program is running, you could stop the pipeline, set a new resolution with a cfg.enable_stream stream configuration instruction and start the pipeline again to apply the new resolution setting.

[MartyG-RealSense]

Hi @VAIBHAVPATEL97 Do you require further assistance with this case, please? Thanks!

[MartyG-RealSense]

Case closed due to no further comments received.