-
随着信息技术的不断发展以及媒体传播方式的不断优化,数字图像作为信息传输的媒介已被人们普遍接受[1].由于数字图像具有较好的直观表达能力,被广泛应用于新闻媒体、航空航天、遥感探测等领域.人们可以通过对数字图像进行分析来获取感兴趣的信息.但由于拍摄环境、传输技术等因素的影响,有时会造成数字图像的损坏,从而影响数字图像的后续处理[2-3].因此,为了能够对损坏图像进行复原,获取其完整的图像信息,近年来,诸多学者对图像修复算法进行了研究.如李爱菊等人[4]提出了一种基于改进Criminisi算法的图像修复方法,该算法能够修复破损图像,但是其最优匹配块搜索策略还不够完善,导致修复后的图像存在块效应;金炜等人[5]针对卫星云图的云图数据破损,提出了一种联合块匹配与稀疏表示的卫星云图修复方法,该方法能对卫星云图进行修复,而且还能克服修复图像过程中出现的块效应,但是该方法中的块匹配计算过程复杂度较高,并且修复图像存在模糊效应与视觉间断的问题;祝轩等人[6]提出了一种基于稀疏分解的图像修复方法,该方法具有良好的修复视觉效果,但是却无法确保修复块的修复顺序,导致复原结果不理想.
对此,本文提出了一种基于梯度变换与最优似然法则的图像修复方法.利用像素点对应的邻域方向特征来构造置信度,改进传统的优先权因子,用于判定优先修补块;并根据像素点的梯度变换,建立修补块尺寸选择模型,自适应调整待修补块的尺寸;通过待修补块与匹配块之间的内积关系以及距离关系,建立最优似然法则,完成图像修复;最后,测试所提修复算法的图像复原质量.
An Image Inpainting Method Based on Gradient Transformation Coupled with Optimal Likelihood Rule
-
摘要: 为了解决当前图像修复算法在待修复图像纹理结构较为丰富时易产生模糊效应以及块效应的问题,提出了一种基于梯度变换与最优似然法则的图像修复算法.首先,利用像素点对应的邻域方向特征来构造置信度,用以形成优先权因子.通过优先权因子对待修补块的优先级进行度量,从而确定最优修补块;然后,根据像素点的梯度变换,构造修补块尺寸选择模型,对修补块的尺寸进行自适应调整;最后,利用修补块与匹配块的内积关系、距离关系,分别构造余弦度量模型、相似度量模型,从而建立最优似然法则,从源区域中搜索最优匹配块,对待修复块进行填充修复.实验结果显示,与其他图像修复算法相比,本文算法具备更高的修复质量,能有效克服阶梯效应以及模糊效应.Abstract: In order to solve such defects in current image inpainting algorithms as blurring effect and blocking artifact induced by taking the repaired block with fixed size as template to search the optimal matching block for image inpainting, an image inpainting method based on gradient transformation coupled with optimal likelihood rule is proposed in this paper. Firstly, the confidence is constructed by using the neighborhood direction feature of pixel points to form the priority factor. And the optimal patch block is determined by using the priority factor to measure the priority of the patch blocks. Then, a repair block size selection model is constructed based on the gradient transformation of the pixel points to adaptively adjust the size of the patch block. Finally, the cosine metric model and similarity measurement model are constructed, respectively, based on the inner product relation and the distance relation between patches and matched blocks to establish the optimal likelihood rule for searching the best matching block from the source region and filling and repairing the repaired blocks. The results of an experiment have shown that compared with the current image inpainting algorithms, this algorithm has better repair quality which can effectively overcome the staircase effect and the blur effect.
-
[1] 王丽丹, 段书凯, 段美涛.忆阻Fourier神经网络在图像复原中的应用[J].西南大学学报(自然科学版), 2014, 36(1): 1-6. doi: http://xbgjxt.swu.edu.cn/jsuns/jsuns/ch/reader/view_abstract.aspx?file_no=2014-01-001&flag=1 [2] 高晓琴, 晏勇, 唐琦.基于可逆数字水印的图像认证新方法[J].西南师范大学学报(自然科学版), 2015, 40(7): 119-123. doi: http://d.wanfangdata.com.cn/Periodical_xnsfdxxb201507021.aspx [3] CASACA W, BOAVENTURA M, DE ALMEIDA M P, et al. Combining Anisotropic Diffusion, Transport Equation and Texture Synthesis for Inpainting Textured Images[J]. Pattern Recognition Letters, 2014, 36: 36-45. doi: 10.1016/j.patrec.2013.08.023 [4] 李爱菊, 钮文良.基于改进Criminisi算法的图像修复[J].计算机工程与应用, 2014, 50(18): 167-170. doi: 10.3778/j.issn.1002-8331.1401-0419 [5] 金炜, 王文龙, 符冉迪, 等.联合块匹配与稀疏表示的卫星云图修复[J].光学精密工程, 2014, 22(7): 1886-1895. doi: http://www.cqvip.com/QK/92835X/201407/661720714.html [6] 祝轩, 张旭峰, 李秋菊, 等.基于稀疏分解的图像修复方法[J].计算机科学, 2016, 43(1): 294-297. doi: 10.11896/j.issn.1002-137X.2016.01.063 [7] 张婷曼.基于小波阈值自适应修正的模糊图像修复算法[J].控制工程, 2015, 22(6): 1166-1170. doi: http://industry.wanfangdata.com.cn/yj/Detail/Periodical?id=Periodical_jczdh201506029 [8] 江平, 张锦.一种结合CDD模型和Criminisi算法的图像修复算法[J].图学学报, 2014, 35(5): 741-746. doi: http://doi.wanfangdata.com.cn/10.3969/j.issn.2095-302X.2014.05.014 [9] doi: http://www.degruyter.com/view/j/jisys.2013.22.issue-3/jisys-2013-0033/jisys-2013-0033.xml?format=INT BIRADAR R L, KOHIR V V. Texture Inpainting Using Covariance in Wavelet Domain[J]. Journal of Intelligent Systems, 2013, 22(3): 299-315. [10] KHELLAH F. Textured Image Denoising Using Dominant Neighborhood Structure[J]. Arabian Journal for Science and Engineering, 2014, 39(5): 3759-3770. doi: 10.1007/s13369-014-1057-z [11] 张琳娜, 赵凤群.基于偏微分方程图像修补BSCB模型的应用[J].电子测试, 2014(16): 22-24. doi: 10.3969/j.issn.1000-8519.2014.16.011 [12] 朱晓临, 陈晓冬, 朱园珠, 等.基于显著结构重构与纹理合成的图像修复算法[J].图学学报, 2014, 35(3): 336-342. doi: http://mall.cnki.net/magazine/Article/GCTX201403003.htm [13] 钱方, 孙涛, 郭劲, 等.基于小波变换的多尺度SSIM算法[J].液晶与显示, 2015, 30(2): 317-325. doi: http://www.cnki.com.cn/Article/CJFDTotal-YJYS201502021.htm [14] 张志伟, 马杰, 夏克文, 等.一种应用于图像修复的非负字典学习算法[J].光电子·激光, 2014, 25(8): 1613-1619. doi: http://www.cqvip.com/QK/92586A/201408/662047015.html [15] doi: https://www.researchgate.net/profile/Venkatachalam_Chandrasekaran2/publication/257471990_Exemplar-based_color_image_inpainting_A_fractional_gradient_function_approach/links/54b9f8e10cf2d11571a4bb25.pdf SAI HAREESH A, CHANDRASEKARAN V. Exemplar-Based Color Image Inpainting: A Fractional Gradient Function Approach[J]. Pattern Analysis and Applications, 2013, 17(2): 389-399. [16] 蔡占川, 姚菲菲, 唐泽圣.基于克里金插值法的图像修复[J].计算机辅助设计与图形学学报, 2013, 25(9): 1281-1287. doi: http://d.wanfangdata.com.cn/Periodical/jsjfzsjytxxxb201309003 [17] 胡文瑾, 刘仲民, 李战明.一种改进的小波域图像修复算法[J].计算机科学, 2014, 41(5): 299-303. doi: 10.11896/j.issn.1002-137X.2014.05.064 [18] 廖斌, 苏涛, 刘斌.基于多尺度分解的k邻域随机查找快速图像修复[J].电子与信息学报, 2015, 37(9): 2097-2102. doi: http://www.cqvip.com/QK/91130A/201509/665908581.html [19] 吕永利, 姜斌, 包建荣.基于像素权值的高效小波图像修复算法[J].信息与控制, 2015, 44(1): 104-109. doi: http://kns.cnki.net/KCMS/detail/detail.aspx?filename=xxyk201501017&dbname=CJFD&dbcode=CJFQ [20] LI K S, WEI Y S, YANG Z, et al. Image Inpainting Algorithm Based on TV Model and Evolutionary Algorithm[J]. Soft Computing, 2016, 20(3): 885-893. doi: 10.1007/s00500-014-1547-7 [21] LIU Y, LIU C J, ZOU H. A New Structure Tensor Based Image Inpainting Algorithm[J]. International Journal of Grid and Utility Computing, 2016, 7(4): 294-303. doi: 10.1504/IJGUC.2016.081015 [22] JIDESH P, BINI A A. A Curvature-Driven Image Inpainting Approach for High-Density Impulse Noise Removal[J]. Arabian Journal for Science and Engineering, 2014, 39(5): 3691-3713. doi: 10.1007/s13369-014-0983-0