YOLO looks at the image once and divides the image up into a grid. It then uses bounding boxes to divide up the grid. YOLO first determines whether the bounding box has an object and then determines the class of object. By incorporating a prefilter on the algorithm, this screens out parts of the images that are not objects and YOLO is then able to dramatically speed up its search.
In this example, after importing our libraries, we set our variables. First, we open yolov3.txt. This file contains the classes of the pretrained library we will be using. Next, we create a random color array to denote our different objects as different colors. Then we import our libraries and set our camera to the first camera on the computer. We then set thresholds and scale images so that the image sizes are something that would be recognizable to the classifier. If we, for example, add a high-resolution image, the classifier might only recognize very small things as objects while...