OBJECT DETECTION AND DISTANCE ESTIMATION USING YOLO ARCHITECTURE

Abstract

Detecting various classes of objects and measuring the distance between camera and objects are implemented in this system. The input of the system is an image or video which are captured by the camera. YOLOv5 architecture is used to detect the objects of input image and calculated the distance between camera and detected objects. If the object is detected, the result will be shown with distance meter values. The major objective of this system is to find the instances of each object in digital photos or real time videos predictions. In this system, the focal length value is employed to calculate the distance between the camera and the object. The focal length is calculated using the triangle formula utilizing the bounding box’s width and height parameters. YOLOv5 object detector provided the height and width values of bounding box. Object detection is also useful in video surveillance and image retrieval systems. In the future, this system can be upgraded as a part of the autonomous vehicles to move automatically in real environment.