Publication

Abstract : The main purpose of this research paper is to study and report how the object detection and identification of multiple objects in an image or a video frameworks in real life. As part of this research, we have clearly presented about image classification, image localization, and image detection. The entire work is carried out with the aid of YOLOv3 (You Only Look Once, Version 3) is an object detection algorithm which helps to detect real-time objects. YOLOv3 gives accurate results. It helps in declaring the localizers, classifiers, and detection. It uses a single neural network to the full image to detect the objects using bounding boxes thereby resulting in accurate outputs. The experiments and results show that YOLOv3 is extremely accurate. This rapidness in detecting objects is helpful in situations where humans can't detect with their naked eye cannot track multiple items simultaneously in an image or video frame.