Approaches to Design Zigzag Drive Detection Model Using Image Processing

摘要

It is required to classify the traffic violation for reducing the rate of road accidents and traffic safety. Driver's behavior is one of the main cause that contribute to increase the traffic accidents, as it leads to degrade the performance of driving pattern. This behaviors need to detected and minimize. The paper presented the model, by using video capturing from road side, the diving pattern can be detected. The models then has tested on multiple videos. Mainly two approaches have presented here. Model developed by Canny edge detection method is giving accuracy of 79.16% and model by centroid & blob method is giving accuracy of 83.33%.