A novel algorithm of IC defect images enhancement based on histogram equalization and IHS transform

Integrated circuit (IC) defective image enhancement is very important for the classification and identification of IC real defects. This paper proposes a new algorithm of IC defect image enhancement. Firstly, histogram equalization is used in the IC defective image, which is to improve the contrast of image. Secondly, the image is transformed from RGB to IHS color space, and then space lightness is exchanged. Finally, the fused image from IHS inversely transforms to RGB to get a new image fusion. Through objective and subjective evaluation, it shows that the algorithm has fantastic effect on enhancement of IC defect images, compared to the HE and MSRCR algorithm that processed separately, and can better describe IC defects details. Experimental results show that the improved algorithm in the brightness, contrast and detail of images have been greatly enhanced, not only to improve the overall visual effect of the image, but also to get the maximum information entropy. So the proposed algorithm is conducive to the detection and defect feature extraction of the defects, and lays a foundation for the real defect classification and recognition.