Research On HE Enhancement And FCM Segmentation Algorithms For Printed Ciruit Board Image

With the rapid development of electronic technology, PCB(Printed Circuit Board) is playing an increasingly important role in people’s daily life.Therefor, how to improve the quality of the PCB has incresasingly been an interest problem.Because of PCB defect detection technology based on image processing has higher defect detection speed, better accuracy and other unique advantages has been attracting the attention of many scholars. With the development of the PCB industrial technology, the development of the PCB image defect detection technology based on image processing technology should be constantly improved.Defect detetion based on image processing technology mainly can be divided into two parts of hardware and software. Design of the software part palys a key role for the detect result; software is the core of defect detection system. In this paper, our main work is PCB image defect detection based on image processing as an example to study the key algorithms of defect detection.On the basis of analyzing the advantages and disadvantages of existing algorithm, aimed at the disadvantages of these algorithms based on actual situation in the industrial production, some improved algorithms have been proposed in this paper. On the Windows 7 operating system, the algorithms are implemented in the MATLAB R2010b(32 bit) environment. Specific tasks of this paper are as follows:In the aspect of image enhancement, the existing image enhancement method based on histogram equalization have been analyed and compared. Based on the idea of histogram segmentation, an improved histogram equalization algorithm has been proposed.The new algorithm can reserve the details of the input image as much as possible, while enhancing input image.In view of the image segmentation, Fuzzy c-means clustering(FCM) image segmentation algorithm has the defects of run time consuming and very sensitive to noise. According to these two questions, based on the combination of histogram, an improved fast FCM algorithm and an improved FCM algorithm can restrain the salt and pepper noise is proposed.We testy them by experiment.The results of the simulations show that the improved algorithms achieve the expected purposes.The improved algorithms have better efficiency and are more suitable for the requirements of the industrial production.