Noise Resilient Thresholding Based on Fuzzy Logic and Non-linear Filtering

This chapter proposes a novel image thresholding technique resilient to noise based on the fuzzy logic and the non-linear filtering. The rationale of the proposed technique is to use adaptive Kuwahara filtering, which essentially suppresses the noise while retaining the image texture information. Finally, an optimized threshold selection process is formalized using the fuzzy logic properties. The performance of the proposed technique is then evaluated by extensive experiments on different types of synthetic, real-world images, and floor plan images. Finally, the superiority of the proposed technique is carried out by detailed comparison with the existing methods.