Automated digital visual inspection with dark-field microscopy
Jorge L. C. Sanz, Fritz Merkle, Kwan Y. Wong · Journal of the Optical Society of America A · 1985
In this paper we deal with the problem of detecting and segmenting objects in textured dark-field digital imagery for automated visual-inspection applications. We first present a technique for correcting optical shading effects in conventional dark-field microscopy. After compensating for possible imperfections in the optical setting we address the problem of segmenting objects (defects) in textured dark-field images. The technique that we will follow is based on a sequential application of local operators, which serves the purpose of clustering the object and the background gray levels. This procedure can be considered an extension of average-thresholding-type techniques. Both algorithms for shading correction and object segmentation have fast implementations in general-purpose image-processing pipeline architectures, and therefore they are appealing to real-time computer vision applications. Computational examples showing the appropriateness of the shading-correction procedure as well as the effectiveness of the segmentation wil be discussed.