为了改善PCB三维CT图像的对比度和动态范围,针对目前的双直方图均衡算法无法适应PCB图像特性的问题,提出一种基于直方图峰值的双直方图均衡算法。利用PCB三维CT图像的灰度直方图中的最大值对应于PCB基板的概率最大的先验信息,提出以直方图峰值所对应的灰度值为二分点对直方图进行二分,在其两侧分别进行直方图均衡增强,最后再合并增强结果。实验结果表明算法对PCB三维CT图像更具有针对性,增强效果更好。
A bi-histogram equalization algorithm based on the maximum histogram coefficient is proposed to improve contrast and dynamic range of PCB three-dimensional CT images and to eliminate the problem that current bi-histogram equalization algorithms cannot adapt to the characteristics of PCB images. By fully applying the prior knowledge that the maximum histogram coefficient corresponds to the laminate of PCB in all probability, the algorithm proposes to select the gray level with maximum histogram coefficient as the bipartition point. Histogram equalizations are respectively carried out on the both sides of this bipartition gray level. The final image enhancement result unites these two equalizations. Experimental results show that the proposed algorithm can obtain better-per-formance for PCB three-dimensional CT images.