This might sound an odd question but is there some way or some example for python that shows how portions or the video stream (sections of the image) can be ignored and not processed?
I ask this because when you process video for multi camera, multi target for example which uses reidentification the person being tracked is lost and reacquired on the far end of the view. This means that a new model or reidentification process is created for this person. However a secondary camera that shows the person in the middle of the view might be more advantageous for reidentifying correctly.
To this end I wanted to know if we can zone areas of the view/image for processing and ignore areas outside this view. This would enable to select only the best most prominent area of the view for processing.
Although Multi Camera Multi Target Python* Demo demo does not provide implementation you are looking for, but in general, demo use OpenCV data types, so you may refer to OpenCV documentation to learn how to extract part of image (AKA Region Of Interest) to feed this part in further processing pipeline instead of original whole image.
Hello Elad Segev,
This thread will no longer be monitored since we have provided a solution. If you need any additional information from Intel, please submit a new question.