Image Inpainting for Object Removal Based on Adaptive Two-Round Search Strategy
Lei Zhang, Minhui Chang · IEEE Access · 2020
In traditional inpainting method for object removal, the SSD (Sum of Squared Differences) is always used to measure the degree of similarity between exemplar patch and target patch. Although the matching rule is simple, there is a risk that the target patch is replaced by an unsuitable exemplar patch, which leads to the mismatch error. Even worse, the error may be constantly accumulated along with the process progresses, finally some unexpected objects will be introduced into target region, and the restored image cannot meet the requirements of human vision. In view of these problems, we propose an inpainting method based on adaptive two-round search strategy. Firstly, we define the DBP (Differences Between Patches) between target patch and exemplar patch, and use it to measure the degree of difference between the two patches. Then, based on SSD and DBP, we adaptively judge whether there is a mismatch error. If the mismatch error occurs, the two-round search strategy is implemented. We define a new matching rule and use it to re-search the exemplar patch. Finally, we use the exemplar patch to restore the target patch. Experimental results demonstrate the effectiveness of our method. It can effectively prevent the occurrence of mismatch error and error accumulation, improve the restoration effect.