Traffic Flow Amount Detection Technique Research Based On Morphology

The traffic flow amount is an important character of ITS(Intelligent Transport System) and real-time, stable traffic flow amount is important basis to judge traffic condition. Comparing with the traditional detection technique, the traffic flow amount detection technique based on video contains the following advantages:abound in information, flexible in setting and low cost, which has been a hot topic home and abroad. It is analyzed that the traffic flow amount detection technique in detail in the paper, which is based on the real traffic condition and combines with character of traffic flow video. In addition, based on morphology image processing technique, it is particularly studied that the repeated detection of cross-road driving automobile and missing detection of parallel driving automobile. It is designed that the dynamic algorithm of road based on virtual detection line and Matlab simulation test are made in the paper, which acquires a good result.It is classified that the detection lines according to the width of automobiles and one detection line stands for one lane, which is based on denoising of morphology. Because the lane is dynamic, the parallel driving automobile can be detected by the algorithm to some degree, which can comparing accurately detect the parallel driving cars and not miss any one, and overcome the drawbacks of traditional detection ways. Meanwhile, if the detection line is wide enough, it can be lowered that the repeated calculation of cross-road driving automobile, which can improve the detection accuracy of traffic flow amount.In addition, it is combined that the character of dynamic object image series in the paper, which is based on the traffic flow amount detection technique. It is used that the absolute value scale to difference image to demonstrate the differences of images. What's more, it is proposed that the new way combing max variance threshold way with gray threshold way in the paper and the algorithm that provides filling for the inanition of the image and horizontal concavity, which provides powerful guarantee for the dynamic algorithm based on the detection line.