印刷电路板光电图像获取与预处理研究

Study of Capturing and Preprocessing of Printed Circuit Board Photoelectric Image

为了获取完整且大面积的印刷电路板(PCB)光电图像,在淡入淡出的自然缝合算法的基础上,提出了一种基于硬件的图像自动拼接新算法,并分析了其基本原理。选择用中值滤波滤除图像中的噪声。针对获取后的PCB光电图像整体偏暗的特点和其直方图中出现灰度冗余的情况,提出了去灰度冗余与灰度变换结合起来的图像增强新算法,并分析了其基本原理。用图像自动拼接算法得到了完整且大面积的PCB光电图像实验结果;用完整图像中的局部小图像从主观与客观两方面对几种图像增强算法的增强效果进行了比较,本图像增强算法的增强效果较好。

On the basis of natural suture algorithm of fade in and fade out,a novel image automatic mosaic algorithm based on hardware is presented in order to capture whole and massive printed circuit board（PCB） photoelectric image,and its basic principle is analyzed.Noise in the PCB photoelectric image is filtered through median filtering. Because of the character of unitary obscuration of the image captured and its gray redundance in image histogram,a novel algorithm of image enhancement that combines removing gray redundance and using gray transform is presented,and its basic principle is analyzed.The whole and massive PCB photoelectric image is captured by the image mosaic algorithm proposed in the experiment.After image preprocessing,the enhancement result of the small part of the whole and massive PCB photoelectric image with our enhancement algorithm is better than that by other enhancement algorithms on the aspects of subjectivity and objectivity.