全局阈值自适应的高亮金属表面缺陷识别新方法

发布时间：2019-04-08 10:09

【摘要】：高反射类零部件在其生产及后期处理过程中,可能会产生划痕、擦伤等表面缺陷,严重影响产品的使用性能和寿命。该类零件表面具有镜面反光特性,易导致检测过程中缺陷目标的漏检、错检。针对这类问题,基于数字图像处理技术,提出一种具备全局阈值自适应调整的高亮表面缺陷识别新方法。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,保护目标边缘信息;其次,以高斯函数的一阶导数构建Canny最优边缘检测器,结合全局阈值最大类间方差法和形态学图像分割法,完成图像分割以及相应阈值的自适应调整,实现对缺陷目标的识别。实验结果验证了算法的有效性及可靠性,能够在排除高光影响的基础上有效地识别缺陷目标,对高亮金属表面缺陷识别具有重要意义。
[Abstract]:In the process of its production and post-treatment, high-reflection components may cause surface defects such as scratches, scratches, and so on, which seriously affect the service performance and service life of the product. The surface of the part has the characteristics of mirror reflection, which is easy to lead to the leak detection and the error detection of the defect target in the detection process. In view of this kind of problem, a new method for identifying surface defects with global threshold self-adaptive adjustment is proposed based on digital image processing technology. The method comprises the following steps of: firstly, constructing a filtering mode for processing an original image by using a spatial domain and a value range information, and protecting the target edge information; secondly, constructing a Canny optimal edge detector with a first derivative of a Gaussian function, combining a global threshold maximum inter-class variance method and a morphological image segmentation method, And the self-adaptive adjustment of the image segmentation and the corresponding threshold is completed to realize the identification of the defect target. The experimental results verify the validity and reliability of the algorithm, can effectively identify the defect target on the basis of removing the high light influence, and has important significance for identifying the defect of the metal surface.
【作者单位】：西安理工大学机械及精密仪器工程学院;西安工程大学;
【基金】：国家自然科学基金(51505359)项目资助
【分类号】：TG84

【相似文献】