Deep Learning for Real-Time Vehicle Counting and Categorization in Traffic Videos

Abstract - Traffic analysis is a problem that megacity itineraries have been dealing with for times. Smarter styles are developed to anatomize business and speed up the process. Business analysis can record the number of vehicles and vehicle classes in an area at a given time. People have been developing similar mechanisms for decades now, but utmost of them involve using detectors to calculate the direction of moving vehicles and identify vehicles to track vehicle figures. Although this system has progressed over time and is truly effective, they're n't budget-friendly. The problem is that similar systems bear periodic conservation and estimation. thus, this design aims to calculate and classify the vehicle rested on vision. The system involves wharf frames from videotape to descry and count vehicles using Gaussian Mixture Model (GMM) background deduction, also classifying vehicles by comparing figure areas with prognosticated values. A significant donation of the paper is the comparison of two type styles. type is done using Contour Comparison( CC) and Bag of Features(BoF) styles. Keywords: - Vehicle counting, Business analysis, Contour Comparison.