摘要
为解决交通事故现场轮胎花纹痕迹的识别问题,研究了适用于事故现场轮胎花纹图像的预处理算法,实现了轮胎花纹图像的平滑、滤波和锐化处理,并提出了基于Canny算法的边缘检测算法,最后基于分水岭算法对轮胎花纹图像进行特征分割,实验结果表明,本文提出的轮胎花纹处理算法能有效地对轮胎痕迹图像进行边缘检测和特征提取,识别结果准确。
In order to solve the problems in recognizing the marks of the tire tread pattern in the traffic accident scene, the prepossessing algorithm suitable for the images of the tire tread pattern in the accident scene was studied. After realizing the purpose of smoothing, filtering and sharpening the images of the tire tread pattern, the paper proposed the edge detection algorithm based on Canny algorithm. Finally, the images of the tire tread pattern were segmented based on the watershed algorithm. The results of the experiment showed that the tire tread pattern processing algorithm proposed in this paper could effectively detect and extract the edge of the images of tire marks and the recognition results were accurate.
出处
《道路交通科学技术》
2019年第2期7-12,共6页
Road Traffic Science & Technology
基金
公安部交通管理科学研究所课题(2018SJA01)。
关键词
交通事故
轮胎花纹
图像处理
边缘检测
特征分割
traffic accident
tire tread pattern
image processing
edge detection
feature segmentation
