Oriented boundary padding for iterative and oriented fringe pattern denoising techniques

Optical interferometric techniques offer non-contact, high accuracy and full filed measurement, which are very attractive in various research and application fields. Fringe patterns are the recorded results of these techniques and often require denoising at the pre-processing step to increase the accuracy and robustness of information retrieval. Among various fringe pattern denoising techniques, iterative and oriented denoising techniques based on partial differential equations in the spatial domain are effective and widely used. However, these techniques introduce errors near boundary areas if traditional image padding methods such as zero padding and symmetric padding are used. Due to a large number of iterations needed in these techniques, the error will flood from the boundary into the inner part of the fringe pattern. Since fringe patterns have a flow-like structure represented by fringe orientation, padding along the fringe orientation helps to reduce the error. An oriented padding method is thus proposed for iterative and oriented fringe pattern denoising techniques that require partial derivative estimations. Simulated fringe patterns are tested and quantitative results are given to demonstrate the performance of the proposed method. Experimental results are also given for verification.