|
|||||
|
|
| 基于结构特征的路面裂纹病害检测算法 | |
| 引用本文： | 王刚,王娟,王德华,张志峰,肖亮.基于结构特征的路面裂纹病害检测算法[J].仪器仪表学报,2010,31(2). |
| 作者姓名： | 王刚 王娟 王德华 张志峰 肖亮 |
| 作者单位： | 1. 鲁东大学电子与电气工程学院,烟台,264025 2. 青岛农业大学理学与信息科学学院,青岛,266109 3. 南京理工大学计算机科学与技术学院,南京,210094 |
| 基金项目： | 国家自然科学基金，山东省教育厅科研项目，鲁东大学科研基金，山东省高校科技计划项目，鲁东大学创新团队建设项目、鲁东大学学科建设经费资助项目 |
| 摘 要： | 提出一种新颖的图像分解算法,用于复杂背景下的路面病害图像,可以获得较为清晰的裂纹病害信息。通过对含有局部线性特征的裂纹图像进行形态分析,建立与结构特征相匹配的基底,利用局部Ridgelet函数作为稀疏表示的基底,图像中的结构信息作为一种组成成分被提取出来,同时运用全变差正则化方法对提取的结构信息加以正则化约束,使得提取结果更加精细化。实验结果表明利用该算法提取的裂纹病害信息符合人眼的观测结果。 |
| 关 键 词： | 结构特征 局部Ridgelet变换 全变差正则化 误差范数 路面裂纹 |
Detection algorithm for pavement cracks based on structure characteristic |
|
| Wang Gang,Wang Juan,Wang Dehua,Zhang Zhifeng,Xiao Liang.Detection algorithm for pavement cracks based on structure characteristic[J].Chinese Journal of Scientific Instrument,2010,31(2). | |
| Authors: | Wang Gang Wang Juan Wang Dehua Zhang Zhifeng Xiao Liang |
| Abstract: | A novel image decomposing algorithm is proposed in this paper, which can obtain clear pavement crack information under complicated background. By means of morphological analysis of the image containing local linear features, the basis that matches the structural features can be set up. The structure information in the image can be drawn as one component by using local Ridgelet as the basement of sparse decomposing. Moreover, total-variation regularization scheme can be utilized to regularly restrict the structure information and the results can be drawn more meticulously. Experiment results show that the crack information drawn by the proposed method is in accordance with human visual system. |
| Keywords: | structure characteristic local Ridgelet transform total-variation regularization error norm pavement crack |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |