关键词: 灰度共生直方图, 局部二值模式, 梯度直方图, 支持向量机

Abstract:

With the rapid development of society and economy，transportation is also developing rapidly，and the problem of urban traffic congestion is becoming increasingly serious ，which not only increase people's travel time and cost，but also cause large economic losses.The main solutions are widening roads ，reducing travel as well as inducing traffic which means effective traffic guidance is inseparable from accurate judgment of traffic conditions. Nowadays,the transportation environment is complicated with complex as well as computationally intensive existing methods of classification and evaluation of transportation status. Traditional methods of obtaining traffic parameters through image processing require separate modeling and model updating for each road segment，while video or the method of image extraction of traffic parameters and feature parameters is simpler and more convenient than traditional methods，and it is easy to maintain.This paper proposes a method for classifying traffic flow or congestion by extracting feature parameters and using support vector machines，and can combine and classify data on different road sections，which effectively improves the efficiency of classification and evaluation of transportation status. Simulation experiments verify the effectiveness of the proposed method.

Key words: gray level co?occurrence histogram, local binary pattern, histogram of oriented gradient, support vector machine