金属表面脏污区域的分割和量化算法研究

摘要

针对目前金属表面缺陷检测技术的研究现状,主要研究金属表面脏污检测与量化算法.结合金属表面的特点,通过对几种分割方法的比较,研究了阈值分割法中的迭代法对图像进行分割,该算法能够最大限度地保留细节信息.另外,研究了一种改进的自适应滤波算法结合像素计数法对金属表面的脏污区域进行量化处理.实验结果证明了所提出方法的有效性.%According to the present research status of defect detection on the metal surface, this article mainly aims at the detection and quantization of smudgy areas on the metal surface. Combining with the metal surface characteristics, based on the comparison of several segmentation methods, the iterative threshold method is studied for image segmentation, which can furthest keep the detail information. In addition, an improved adaptive filtering method and the pixel counting method are used to quantify the smudgy areas on the metal surface, the experimental results show the effectiveness of the method.