Detection method based on a coupled illumination correction algorithm for the detection of surface defects in ZrO

To address the problem of feature texture loss in defective images of zirconia bearing balls on unbalanced illumination backgrounds, a coupled algorithm of the local gamma function transform-block interpolation balancing equation is designed. A combination of a light reflection imaging model and a three-dimensional Gaussian surround function is used to extract the illumination estimate of the original image. Subsequent illumination balance correction is performed in the value color space. The contrast between the edges of the feature texture and the background is analyzed to construct a locally adaptive gamma function transformation model. We establish local block interpolation balance equations to eliminate the effects caused by illumination that is too high or too low. The experimental results show that the information entropy of the images after illumination correction changes in the range of 5% or less. The standard deviation and mean gradient can be increased by 15.21% and 14.78%, respectively.