Project 9

DeepSORT is a computer vision tracking algorithm for tracking objects while assigning an ID to each object. DeepSORT stands for Deep Simple Online RealTime Tracking.

The first step is to object detectors to detect the object in the frame that they are to track. These detections are then passed on to the next step. In estimation, we propagate the detections from the current frame to the next frame which is estimating the position of the target in the next frame using a constant velocity model. when the detection is associated with the target, the detected is associated with the target, the detected bounding box is used to update the target state where the velocity components are optimally solved.

We have the target bounding box and the detected bounding box. So a cost matrix is computed as the intersection-over-union distance between each detection and all predicted bounding boxes from existing targets.

The steps used to track the objects are:

• First, we import the necessary modules.
• Then we set the bounding boxes by using the cv2.rectangle function where we bound each object.
• Then we track the id by creating the empty arrays previous point and next points
• first, we calculate the distance of points if the distance is less than 20 we assume it is the same object
• else increase the value of id.
• Then we remove the unwanted id ie already known id values.