Evaluating Spherical Gaussian Fuzzy Sets in Image Enhancement

Fuzzification techniques play a pivotal role in enhancing image processing tasks by effectively modeling uncertainty and imprecision inherent in visual data. This study presents a new approach using image fuzzification and subsequent defuzzification. We define the image representation using its fuzzy components and perform the enhancement with cutting-edge spherical fuzzy sets. We compare the performance of enhancement with Gaussian (Type-1), Interval Type-2 Gaussian sets using contrast, edge strength and entropy scores. We systematically fuzzify images and analyze the effectiveness of each membership function in capturing image ambiguity as a complewx function. Our results highlight the trade-offs between computational complexity and fuzzification fidelity, with Interval Type-2 and spherical fuzzy sets demonstrating superior uncertainty handling and detail preservation compared to conventional Gaussian fuzzification. This work underscores the potential of advanced fuzzy sets in improving image processing pipelines, particularly for applications requiring robust handling of ambiguity.