摘要： 印刷电路板(PCB)在现代工业中发挥着重要作用,通过锥束CT对PCB进行三维无损检测具有重要意义。但是在CT成像过程中,存在金属伪影、射束硬化等诸多影响图像质量的因素,导致低对比度PCB三维CT图像的出现。针对PCB三维CT图像中由于对比度低造成电路信息难以分辨的问题,提出一种基于变权直方图的图像增强算法。在基于直方图生成累积概率密度函数时,通过变权累积求和的方式,把灰度值大小、灰度值概率以及灰度值区间附近的概率分布引入到变权策略中,生成一个兼顾提高对比度和扩展灰度范围的灰度映射曲线,实现低对比度图像的增强,提高PCB图像中物体的可分辨性。实验结果表明,该算法取得了较好的增强效果和图像质量。

关键词: 图像增强, 直方图均衡, 低对比度图像, 三维图像, 印刷电路板, CT图像

Abstract: Printed Circuit Board(PCB) is playing an important role in the modern industry,while Cone Beam Computed Tomography(CBCT) is an effective way for PCB imaging and detection.The problem of low contrast of PCB always happens because of the artifacts introduced during the CT imaging.Aiming at the problem of enhancing the three-dimensional PCB low contrast CT image,this paper proposes a method based on variable weight cumulating histogram equalization.The value and probability of the gray and the probability distribution of the span close to the gay are introduced in the strategy of altering the weight when constructing the cumulative density function.A gray transfer curve considering both enhancement effect and gray level expand is produced to promote the metal in the low contrast PCB image,thus enhances the image.Experimental results show that this algorithm can obtain better enhancement effect and image quality.

Key words: image enhancement, histogram equalization, low contrast image, three-dimensional image, Printed Circuit Board(PCB), CT image

中图分类号: 

• TP301.6