Abstract: In older to solve the problem of the deviation accumulation caused by the restoring process of exemplar-based image inpainting algorithm, and to improve the accuracy of the algorithm, this paper focused on the revision of the priority computation formula, by leading an α factor to enhance the proportion of data. Therefore, the algorithm is more sensitive to the detail of image texture during the inpainting procedure. The average energy of texture feature was quantified by the wavelet coefficients. Thus the α factor was made to be self-adaptive so that the images with different energy could be inpainted adaptively through appropriate strategies. It is proved that the image quality can be enhanced effectively by this improved exemplar-based image inpainting algorithm, and the optical connectivity requirements of human beings can be satisfied by the inpainted effects.

Key words: image inpainting, self-adaptive, wavelet transform, priority, texture feature

摘要： 为了解决基于样本图像修复算法对纹理部分的修复易产生误差累计的问题，提高图像修复的准确性，对优先值计算公式进行了修正，通过引入调节因子α调整填充边缘优先级顺序，使算法在修复过程中对图像纹理细节的部分较为敏感；利用图像的小波系数估计图像的平均细节能量值，自适应地调节α因子，从而实现对不同纹理程度的图像自适当地调整修复策略，并通过实验证明了算法的有效性。

关键词: 图像修复, 自适应, 小波变换, 优先级, 纹理特征