I just want to save the picture of the detected box,how to modify？

[yancccc]

hi, draw_detections_v3(im, dets, nboxes, thresh, names, alphabet, l.classes, ext_output); // save_image(im, "predictions"); save_image(im, outputfilename_without_ext);

I just want to save the picture of the detected box,how to modify？

[vincentgong7]

Hi,

Do you mean the picture of the objects that characterised by the bounding box? If so, I suggest:

1. using Yolo to detect objects from images.
2. save the bounding box information to json.
3. write another python or Java program to Crop out those bounding box pictures from each image according to the bounding box coordinates in the json file.

How do you think about this?

BR, Vincent

[yancccc]

I don't mean to crop out the picture of the bounding box, Only save the picture of detected things, ignore the picture of nothing detected

[vincentgong7]

It's even easier. According to the exported text file(text file or json), it is easy to know which images contain identified objects and which do not have. Then you just need to copy those images contain detected images to one folder.