Traffic Density Recognition Based on Image Global Texture Feature

摘要

Traffic state recognitions can provide a strategic support for control and management of urban traffic, which is crucial to easetraffic congestion, reduce road accidents, and ensure road traffic efficiency. This paper proposes an effective traffic densityestimation method based on image processing. In the beginning, a whole image is divided into several cells, and then a region ofinterest (ROI) is extracted based on calculating varieties of pixel values in a temporal sequence of each cell. Then a texture featuredescriptor, a histogram of multi-scale block local binary pattern (HMBLBP) is proposed for local feature representation. TheHMBLBP of all cells in the ROI are concatenated as a global feature. Furthermore, principle component analysis is performed fordimensionality reduction to save computational cost. At last, the method proposed is tested with two datasets captured from realworldtraffic scenarios. By using the support vector machine (SVM) classifier, traffic states are classified into heavy, medium andlight densities. Reliable performances are shown in the experimental tests.