Open Smellegg opened 4 years ago
Update:
So I have been messing around with image.c around draw_detections.
I attempted many things but so far only managed to make it print woof for every object detected be that dog or not. I think problem is me not knowing is a object detected known as a class[j] name or dets.
I am wondering if it is easier to get the detector to produce a text file instead of predictions.jpg and then run my own script on that predictions.txt instead.
There must be an easier way!
Hi all,
I have been playing with Yolov3 for a while now, training my own detector, and just messing around with the capabilities.
I am trying to figure out how to get the output of the detection to then trigger something else, but I am not sure where to put my code. or how to do it.
I am trying just simple things like
if dog detected print('woof') or if animal detected activate servo
I have seen things about possibly adding a few lines of code to files already in darknet or is it better to write my own script? Any advise greatly appreciated!