Image Inpainting Algorithm Based On Flatness Detection And Contour Similarity

With the rapid development of technology and the advent of the 5G digital era,digital images have become one of the most common sources of information in daily lives.However,in the process of digital image formation,storage and transmission,it is easy to lose information,resulting in digital image damage.Digital image inpainting technology is a method to fill the missing information based on the known information in the image.Therefore,it is necessary to study image inpainting technology to solve the problem of image damage.At present,the exemplar-based inpainting algorithm has become one of the mainstream algorithms in the field of image inpainting due to its significant improvement in the inpainting effect and efficiency.However,it is still a great challenge to fill the image with large damaged area and complex background.Based on the above problems,an image inpainting algorithm based on flatness detection and contour similarity is proposed.The specific work is as follows:1)In this paper,an image inpainting algorithm based on flatness detection and contour similarity is proposed.First,the algorithm introduces a flatness detection factor based on the structural information around the target patch,and a new priority model based on flatness detection is propesed.This model can effectively distinguish texture and structure information and obtain more reliable inpainting order.Then,combining the non-local features of images,the similarity criterion based on contour feature is used to measure the differences between samples,and to find multiple similar patches of the target patch.Finally,a weighted fusion scheme based on singular value difference is used to aggregate all similar patches to reconstruct images.The experimental results show that the proposed algorithm in this paper can effectively avoid the smoothing effect,and can distinguish the differences between samples well and maintain the structural coherence and clear texture for the large areas damaged images,so as to obtain a good inpainting effect.2)In order to solve the problem of high computational complexity caused by the introduction of contour features to measure the differences between samples and uses global search to find the best similar patch in the inpainting algorithm based on flatness detection and contour similarity,this paper proposes an image inpainting algorithm based on biogeography-based optimization algorithm(BBO).Combined with the proposed based on flatness detection and contour similarity inpainting algorithm,this algorithm integrates BBO algorithm into the process of patch matching,and used migration and mutation operations of BBO to search for similar sample patches,which avoids the defect of global search for similar sample patches.Experimental results show that the algorithm can reduce the computational complexity and improve the efficiency of the algorithm while maintaining the structural coherence of the inpainting results.