Image quality assessment.

Museum collection images are invaluable for preserving cultural heritage and studying history. However, these images often lack quality and clarity. This study introduces a novel museum collection image enhancement technique based on fuzzy set theory. The proposed approach comprehensively addresses the complexity and uniqueness of museum collection images to improve their quality significantly. The methodology steps start with preprocessing and graying out the images, eliminating noise through segmentation and smoothing processes. A convolutional neural network (CNN) is utilized to extract image features and apply adaptive histogram equalization for enhancement. A distinctive aspect of our method is transforming image grayscale levels into fuzzy sets. We analyze the similarity between the fuzzy sets before and after enhancement using the cosine similarity algorithm, allowing us to reconstruct the processed images with targeted similarity. Testing our approach on 100 museum collection images, we found that the average contrast of the collection images improved significantly. Specifically, the average contrast for our fuzzy set-based enhancement was 0.91, compared to 0.81 and 0.79 for histogram equalization and wavelet transform methods, respectively. Our research showcases a museum collection image enhancement technology based on fuzzy set theory that effectively enhances image fidelity and clarity, improving the overall quality of museum collection images. Our work underscores the importance of ongoing research in this area to unlock the full potential of museum collection image enhancement technology.