摘要
提出一种基于统计特性的路面裂纹提取方法,用于提取在复杂背景下高速路面的不同裂纹.将路面图像按实际情况分成子块图像,并用各子块图像的灰度均值获取估计背景的平衡参数,结合阈值的迭代选取和Top-Hat变换.大量不同背景图片的处理结果表明,该方法可较好地提取路面裂纹信息.
A method for surface crack extraction based on statistical properties is put forward, which may extract the different cracks of pavement images in complex background. A pavement surface image is divided into sub-images according to actual situation, and their average intensity is used to estimate the balance coefficient of the background. Combining iterative threshold selection and Top-Hat transformation, quantities of different background of images are processed. The experiments show that a good result for extracting pavement crack information can be obtained.
出处
《测试技术学报》
2010年第2期170-174,共5页
Journal of Test and Measurement Technology
基金
山西省高校科技开发基金资助项目(200713013)
中北大学2008年校青年科学基金资助项目
关键词
统计特性
裂纹提取
路面破损检测
图像增强
特征提取
statistical properties
crack extraction
pavement damage detection
image enhancement feature extraction
