A study on improvement of recognition accuracy by applying machine learning algorithms to the vision-based traffic condition analysis system

摘要

This paper proposes a method of applying a machine learning algorithm in order to improve the recognition rate of the video-based traffic information system. After applying the Error Backpropagation learning neural network algorithm to the traffic information it will be used for image recognition results. The training data is generated from the traffic information system, the noise of the generated data is removed by Gaussian smoothing. In this paper, we develop a machine learning based Traffic Condition analysis system was able to get an improved recognition rate than conventional vision-based system.