Comparison of image processing algorithms for estimating parameters from Newton’s rings

As a classic interference fringe pattern, Newton’s rings have wide applications in optical interferometry, especially in measuring physical parameters such as the ring center and curvature radius. Consequently, various image processing methods have been proposed to effectively analyze and process Newton’s ring images. This paper investigates various approaches for the analysis of Newton’s rings and categorizes them into four main groups: spatial domain methods, Fourier transform domain methods, fractional Fourier transform domain methods, and deep learning methods. The basic principles of each method for analyzing Newton’s rings are first discussed, followed by a detailed analysis of their advantages and disadvantages. The performance of these methods was evaluated by comparing the experimental results for both simulated and real interferometric images. Through these analyses, this study provides valuable theoretical guidance for selecting and optimizing Newton’s ring image processing methods and offers insightful recommendations for research and practical applications in this field.