Single Shot Detector Algorithms for Hardware Design on FPGAs for Fast Object Detection

Object identification and analysis take place throughout the process of detecting YOLOv2-based objects. It entails determining the path that an item takes through space and time by identifying that object's location in each object. A YOLOv2-based object detection algorithm is developed by using FPGA in this work. The detection of objects that are in motion is a difficult job. Recent research has shown that convolutional neural networks, often known as CNNs, perform better than traditional methods, and a number of different object detectors that are based on CNNs have been developed. In order to implement the proposed model, the CNN algorithm is employed to distinguish individual items against a backdrop image for reference. MATLAB and an FPGA board are used to carry out the experiment.