Publication: Enhancement of defect regions in x-ray images using histogram equalization based contrast enhancement methods
Loading...
Date
2022-07-01
Authors
Chew, Ken Joo
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Nowadays, for the production of all sorts of PCB, the production for the pads or the exposed region on the circuit board that the components are soldered on are done by machine. Having voids on the single pad is one of the major issues for the production line. For void inspection for the pads, a three-dimensional X-ray image of the PCB will be captured and if there is any void on the pad of the PCB, it will be inspected to determine whether the void is significant enough to be deem a reject or within tolerance. For manual labor inspection, the inspection process is highly reliant to the experience and skill of the manual inspector which is also time consuming. The purpose of this research is to design histogram equalization-based contrast enhancement methods to enhance the defect images and validate the proposed contrast enhancement method using real industry defect images. There are three stages used in the algorithm which are image acquisition, image pre-processing and image processing. For the image acquisition stage, X-ray images obtained are cropped to 185 × 230 and are then loaded into system to undergo system enhancement. Three HE-based contrast enhancement methods, GHE, AHE and CLAHE are then applied. Finally, at the image processing stage, the enhanced cropped X-ray images will undergo image segmentation using the thresholding method. CLAHE works the best when enhancing the X-ray images and the void regions are visible. The thresholding method does not work well for segmenting the void region from the pad region. Only the segmented AHE enhance cropped X-ray image has visible results. With further research, the proposed system could be further enhanced in order to provide better segmentation performance.