摘要
结合数学形态学运算方法,对于板材缺陷图像的提取分割,提出了基于数学形态学的分割方法。利用图像各点间的像素值差异,对板材像素点求取无缺陷标准值,将该值与待测板材进行对应像素点对比,然后将差值求和取平均得到分割阈值;对图像中所有大于该阈值的点进行分割提取,得到分割图像。用数学形态学方法对噪声较大的图像进一步处理,最终得到较为理想的分割图像。
Combined with mathematical morphology method, we propose a segmentation method based on mathematical morphology for extracting plate defects. The main method is to use the image pixel values of the difference between each point for solving the non-de- fective standard. Compared with non-defective plates and the test plates, we take the sum of the difference between corresponding points to average, and we call this average is threshold. All the image point is larger than the threshold segmentation point, we take all the point for our picture. Then using mathematical morphology for larger image noise further processing, we can take zbe picture at last.
出处
《福建林业科技》
2015年第2期73-76,共4页
Journal of Fujian Forestry Science and Technology
基金
黑龙江省自然科学基金(基于计算机视觉的木质板材颜色分类和缺陷检测研究
C201414)
关键词
木材缺陷
数学形态学
图像分割
wood defects
mathematical morphology
timber detection
