Evaluation of Contact Lens Data Acquisition Approaches using Enhancement Techniques

Abstract

Contact lenses can be helpful to improve the quality of human life. The inspection process plays a big role to produce good quality contact lens products. However, there is a challenge to detecting the defects in contact lenses during the production line. The transparent type of silicone hydrogel contact lens is one of the most difficult to detect the defects inside it. The primary purpose of this paper is to examine the differences in quality images between four different data acquisition approaches based on two image enhancement techniques, Gaussian blurring and Contrast Limited Adaptive Histogram Equalization (CLAHE). Acquiring a clear and good-quality image, required a specific experimental setup which consists of a high-resolution camera lens and also the right position of the camera stand and camera angle. Based on performance metrics for both enhancement techniques, Approach 2 showed better performance compared to other approaches when the result from Gaussian blurring showed the highest value of PSNR (29.02321), lowest values of MSE (81.42533), and lowest value AMBE (-0.55510). While for the CLAHE method, the result showed the highest value of PSNR (28.50377), the lowest value of MSE (91.77044), and the lowest value of AMBE (-0.05532). This proves that approach 2 provides a better quality image due to less noise.