基于稀疏逼近的图像编码算法分析与改进  

Analysis and Improvement of Image Coding Algorithm Based on Sparse Approximation

作　　者：吕艳琳[1] 陶玉婷 张燕[1] LYU Yan-lin;TAO Yu-ting;ZHANG Yan(Jinling Institute of Technology, Nanjing 211169, China)

机构地区：[1]金陵科技学院软件工程学院,江苏南京211169

出　　处：《金陵科技学院学报》2020年第4期18-21,共4页Journal of Jinling Institute of Technology

基　　金：教育部产学合作协同育人项目。

摘　　要：在图像表示领域,矩阵分解法近年来得到了越来越多的关注。然而,该类算法大多属于无监督学习方法,故难以有效捕捉数据中固有的几何结构。为此,提出了一种综合应用矩阵分解和稀疏表示的稀疏逼近编码新方法,通过综合应用矩阵分解与稀疏约束构建矩阵优化模型,并通过实验进行验证。实验结果表明:该方法通过稀疏表示,可有效捕捉图像空间的内在几何结构,在识别语义结构方面具有一定的优越性,能更有效地进行图像表示和编码。In the field of image representation,matrix decomposition has received more and more attention in recent years.However,as an unsupervised learning method,it cannot effectively capture the inherent geometric structure of the data.Therefore,this paper proposes a new method of sparse approximation coding that comprehensively applies matrix decomposition and sparse representation.A matrix optimization model is constructed by applying matrix decomposition combined with sparse constraints,and then the corresponding code is written to verify the proposed model.The experimental results of image clustering show that the new method can effectively capture the inherent geometric structure of the image space through sparse representation.It has clear advantages in recognizing semantic structure,and it is more effective for image representation and coding.

关 键 词：图像表示 图像编码 稀疏逼近编码 矩阵分解 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]