Comparative Analysis for Vehicle Detection and Counting System Using Machine Learning

Comparative Analysis for Vehicle Detection and Counting System Using Machine Learning

Darshan Rajput, Tanishq Singhai, Dev Bhadauria, Prof. Ms. Sonia Relan

Abstract : This paper describes a system for detecting and counting the number of vehicles on roads and highways using multiple techniques for extracting image or video features. Globally, the system merely requires a video clip captured by a multitude of static cameras mounted on roads and highways. The ability to recognise and identify a car in a video has proven to be the most difficult and presents numerous difficulties. Some issues may be associated with image processing, such as determining the number of vehicles in an image without human intervention. Where some vehicles can be excluded temporarily or permanently by other vehicles, making detection challenging or impossible. The four phases of the system described are: 1. The variable background subtraction 2. picture segmentation 3. vehicle detection a count of vehicles The need for this in vehicle detection and counting is particularly beneficial for traffic surveillance and many other purposes. Important system capabilities include the ability to identify and count vehicles. The capacity to demonstrate that the system will deliver output with the highest degree of accuracy.

Keywords : Traffic Surveillance, Feature Extraction, Background Subtraction, Segmentation, Detection, Counting, Computer Vision, and Computer Vision.