DESIGN AND IMPLEMENTATION OF AN OBJECT DETECTION SYSTEM USING YOLO

ABSTRACT

This project work on Design and Implementation of an Object Detection System uses the YOLO frame work deep learning algorithm for real-time object tracking and detection. In computer science, object detection; a computer vision technique involves the detection and tracking of object like digital images and videos. In this work, object detection system using YOLO was implemented. The work discusses YOLO concisely and presented the immense benefits of this technology The theory of this deep learning technology as it concerns its literature were discussed and also, using a small custom training data size, the practical implementation of the YOLO object detection system was shown. Lastly, for this work, in order to place values on the computation time, scope of detection and accuracy between our YOLO object detection algorithm and single shot object detector, a single shot detection (SSD) algorithm was likewise written and tested with same data with the YOLO algorithm and the results presented.