How to make YOLO algorithm run detection on images automatically and saves them in a separate folder?

Question

I was working on this project here, and to detect pictures I have to run:

cd darknet
make

then the triggering command line:

./darknet detect cfg/yolov3.cfg yolov3.weights data/dog.jpg

So, my question is, How can I make it run automatically once I throw a picture inside the data folder? So, what I want is, whenever my Motion camera detector detects a motion, it will save it in the darknet/data then the command line will do the analysis or detection (image recognition) and saves it in some other folder!

How possible is that?!

what you want to do is take one image, then another, then compare the images for changes. Motion can do this, as part of the configuration of the motion server. what you can do, probably from the set up scripts for motion-server, is to then run your yolo script command. — j0h, Jan 01 '19 at 05:29

score 1 · Accepted Answer · answered Mar 06 '20 at 08:40

Not sure if you still need help, but there's already a command for you to capture the image from the network camera.

./darknet detect demo cfg/yolov3.cfg yolov3.weights (fill your rtsp address here)

In the source code "image.c" you can copy the following line in a commented section of this function void draw_detections_v3(...):

save_image(cropped_im, image_name);

and modify "cropped_im" to "im" to save the whole screen of each frame (you may need a counter to change the saved names) to your local folder.

How to make YOLO algorithm run detection on images automatically and saves them in a separate folder?

1 Answers1