| 工件表面低对比度缺陷快速准确识别方法 |
| |
| 引用本文： | 邵伟,彭鹏,周阿维,赵迪.工件表面低对比度缺陷快速准确识别方法[J].计量学报,2019,40(5):793-797. |
| |
| 作者姓名： | 邵伟 彭鹏 周阿维 赵迪 |
| |
| 作者单位： | 西安理工大学,陕西西安,710048;西安工程大学,陕西西安,710048 |
| |
| 基金项目： | 国家自然科学基金( 51775433);陕西省自然科学基础研究计划资助项目(2018JQ5180) |
| |
| 摘 要： | 针对工件表面低对比度缺陷,提出一种基于图像网格的灰度统计分析识别新方法,首先构造利用空域和值域信息的滤波方式,对原始图像进行处理;然后对图像进行网格划分,并对网格图像的灰度值熵统计分析,实现对缺陷目标的识别。实验结果表明:对低对比度缺陷识别准确率达到97.1%。
|
| 关 键 词： | 计量学 表面缺陷 低对比度 灰度统计分析识别 边界模糊 |
| 收稿时间： | 2018-12-20 |
Fast and Accurate Recognition of Low-contrast Defects Workpiece Surface |
| |
| SHAO Wei,PENG Peng,ZHOU A-wei,ZHAO Di.Fast and Accurate Recognition of Low-contrast Defects Workpiece Surface[J].Acta Metrologica Sinica,2019,40(5):793-797. |
| |
| Authors: | SHAO Wei PENG Peng ZHOU A-wei ZHAO Di |
| |
| Affiliation: | 1. Xi'an University of Technology, Xi'an, Shaanxi 710048, China
2. Xi'an Polytechnic University, Xi'an, Shaanxi 710048, China |
| |
| Abstract: | At the low contrast defects of the workpiece surface, a new method for grayscale statistical analysis and recognition based on image grid is proposed. First, construct the filtering method using the spatial and range information to process the original image; then carries on the grid division of the image, and the gray value of the image of grid entropy statistical analysis, to achieve recognition of defect targets. The experimental results show that the accuracy rate of low contrast defect recognition reaches 97.1%. |
| |
| Keywords: | metrology surface defect low-contrast grayscale statistical analysis and recognition boundary blur |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《计量学报》浏览原始摘要信息 |
|
点击此处可从《计量学报》下载全文 |
|
