Author(s)

Shuai Liang

Corresponding Author

Shuai Liang

Abstract

The use of traces left behind by traffic accidents can effectively help identify accidental escape vehicles. In the event of a traffic accident, the scene of the accident will generally leave traces of the tires of the vehicle. Inspect and analyze the tire traces to obtain information about the vehicle and help identify and identify the vehicle. Tire pattern image retrieval is an important means to obtain information for solving crimes in traffic accident handling and criminal case detection. Therefore, based on the research of the related literatures in the field of tire pattern image retrieval in recent years, this paper proposes a tire feature comparison system based on multi-level SVM for quick query of the vehicle. Using the database experiment, the comprehensive recognition rate of 99.5% is obtained. High, has great application potential.

Keywords

Tire pattern; cooking vehicle; image retrieval; SVM