ABSTRACT
This chapter focuses on the next important step in building machine vision solutions:, which is digital image processing. Digital image processing is a vast area that covers encompasses many methods, techniques, and algorithms. The choice of methodology depends on the processing requirements. Certain of tThe methods commonly used in machine vision systems for industrial quality control applications in the Industry have beenare explained. Digital image processing has beenis discussed under the broad headings outlined in Chapter 2, namely, Ppre-processing, Ssegmentation, and Oobject Ddetection / Rrecognition. Pre-processing is carried out to improve the quality of the images and emphasize the features required for quality inspection. The chapter explains Ssome of the key operations, such as, Ffiltering, Sscaling / Ssub-Ssampling, and Hhistogram generation have been explained. Image segmentation aims to simplify image processing by portioning the an image into several meaningful parts. Each of these parts can be analyzed and interpreted separately. Thresholding Thresholding-based segmentation, Edge edge-based segmentation, and Region region-based segmentation have beenare described in detail. Object recognition is a challenging task and the various factors that contribute to its difficulty have been discussed,. as are Ttemplate Mmatching and Bblob Aanalysis techniques have been discussed.
