Object Detection and Crop

Uses Darknet's YOLOv3's object detection model (with pre-trained weights) to detect individual objects (persons) in an image, and crops each detection to a file.

TODO:

Output the location of each detection an image in JSON format and file.