For a project I need some special low latency object tracking. I have the code for this available in C++ (for some other hardware platform).
I want to process camera lines as soon as they come in. I do not want to wait for a complete picture. So I would like to have the hardware to read the camera and write the lines in memory using DMA. And parallel on this run the detection software on the latest received lines.
Can I do this with the OpenMV H7 software environment?
Or do I need to write the software from the ground up somehow?
Any advice is welcome!