The Sedan Length Detection Algorithm Based on the Tail Features

Abstract

Extracting effectively the vehicle length is beneficial to classification of vehicle in video traffic detection. In the traditional video traffic detection, vehicles can be approximately regarded as a rectangle to extract the vehicle length. However, it is not accurate enough to use rectangle method to extract the length of sedan. In this paper, based on the tail features of the sedan, the mathematical algorithm of trigonometric function and similar triangles were applied to calculate the sedan length. And then, we discuss the computing method of sedan length detected by video from the different camera angles and detection line locations, and analyze the influence on the accuracy of the measuring sedan length. At last, a study is conducted for example analysis, and the result shows that the approach of extracting sedan length is feasible. Compared with the traditional algorithm, the sedan length accuracy is improved by 3.18% using the algorithm in the paper. Thus, the algorithm is of great significance to improve the accuracy of measuring the sedan length.