Real - Time Object Detection using YOLO Model

Abstract

Object Detection is a PC vision procedure that attempts to recognize and find objects inside a picture or video or in real-time through webcam. In particular, object discovery draws bouncing boxes around these distinguished items, which permit us to find where said objects are in (or the way that they travel through) a given scene. Object location is normally mistaken for picture acknowledgement, so before we continue, it’s vital that we explain the differentiations between them. We are using highly accurate object detection-algorithms and methods such as Region-Based Convolution Neural Network (R-CNN), Fast-RCNN, Faster-RCNN and fast yet highly accurate ones like Single Shot MultiBox Detector(SSD) and You only look once (YOLO). Using the above algorithms and methods, based on the deep learning which is also a part of machine learning that require a lot of frameworks of mathematical and deep learning. Therefore understanding the frameworks by using dependencies such as Tensorflow, OpenCV, cv2, yolov3, pygame etc.