Efficient image inpainting approach using fractional order and optimization algorithm

Image inpainting plays a vital role in restoring damaged images and removing undesired objects; however, exemplar-based inpainting methods often suffer from patch priority degradation and inaccurate patch matching, leading to visual artifacts. To address these limitations, this paper proposes a fractional-order gradient–based exemplar inpainting framework with an improved patch priority computation and optimized patch selection strategy. A regularization factor is incorporated into the priority function to mitigate the priority dropping effect, while a hybrid similarity measure combining Sum of Absolute Differences (SAD) and Mean Squared Differences (MSD) is employed to improve exemplar patch selection accuracy. The Barnacles Mating Optimization (BMO) algorithm is used to determine the optimal fractional order and regularization parameter, and the inpainting process is carried out using the Caputo fractional derivative. Experimental validation on datasets containing both smooth and textured images demonstrates the effectiveness of the proposed approach. Quantitative evaluation shows that the proposed method achieves average improvements of 9.41% in PSNR and 3.08% in SSIM compared with existing state-of-the-art methods. These results confirm that the proposed framework provides improved structural consistency and visual quality in image inpainting.