✅作者简介：热爱科研的Matlab仿真开发者，修心和技术同步精进，matlab项目合作可私信。
🍎个人主页：Matlab科研工作室
🍊个人信条：格物致知。
更多Matlab仿真内容点击👇
智能优化算法 神经网络预测 雷达通信 无线传感器
信号处理 图像处理 路径规划 元胞自动机 无人机 电力系统
⛄ 内容介绍
描述了一种小波编码图像的分割和分析算法。该算法构成了图像后处理方案的一部分，该方案可以成功地恢复遭受模糊伪影的压缩图像中的纹理。该算法包括提取纹理、强度（或颜色）和空间特征。使用k均值算法的变体来有效地分割大图像。分析阶段使用基于规则的启发式方法将片段分类为可用于恢复它们的潜在伪影或相邻纹理。这种新颖的图像后处理方法需要最少的用户交互，并且可以成功地利用压缩图像中的纹理级别相关性。
⛄ 部分代码
function [dog1, dog2, doog21, doog22, doog23, doog24, doog25, doog26] = doog(hsize);
% Difference of Offset Gaussian and Difference of Gaussian Filters
% for Texture Segmentation.
% Based on work by Jitendra Malik and Pietro Perona, "Preattentive texture
% discrimination with early vision mechanisms," J. Optical Soc. America,
% Vol 7, No 5, May 1990.
% (c) 2003 by Rajas Sambhare
% ECE 738 - Final Project
% Spring 2003
% Common parameters
g = 'gaussian'; % Filter type
% Filter 1. Difference of Gaussian 2
sigmaratio = hsize/8*[0.62, 1, 1.6];
sumratio = -[1, -2, 1];
dog2 = sumratio(1)*fspecial(g, hsize, sigmaratio(1)) + ...
sumratio(2)*fspecial(g, hsize, sigmaratio(2)) + ...
sumratio(3)*fspecial(g, hsize, sigmaratio(3));
% Filter 2. Difference of Gaussian 1
sigmaratio = hsize/8*[0.71, 1, 1.14];
sumratio = [1, -1];
dog1 = sumratio(1)*fspecial(g, hsize, sigmaratio(1)) + ...
sumratio(2)*fspecial(g, hsize, sigmaratio(3));
% Remaining filters
aspect = 3; % Fixed aspect ratio of remaining filters.
% Filter 3.
sigmaratio = [1, 1*aspect];
sumratio = [-1, 2, -1];
doog21 = sumratio(1)*gaussian(0, sigmaratio(1), sigmaratio(1), sigmaratio(2), hsize) + ...
sumratio(2)*gaussian(0, 0, sigmaratio(1), sigmaratio(2), hsize) + ...
sumratio(1)*gaussian(0, -sigmaratio(1), sigmaratio(1), sigmaratio(2), hsize);
% Filter 4. Transpose filter 3 to get filter 4
doog22 = doog21';
% Filter 5, 6, 7, 8. Rotate filter 2 (and flip) %%%TODO renumber doogs%%%
doog23 = imrotate(doog21, -30, 'bicubic', 'crop');
doog24 = imrotate(doog21, -60, 'bicubic', 'crop');
doog25 = flipud(doog23);
doog26 = flipud(doog24);
% Display the filters
% figure,
% subplot(3,3,1), mesh(dog2), title('1');
% subplot(3,3,2), mesh(dog1), title('2');
% subplot(3,3,4), mesh(doog21), title('3');
% subplot(3,3,5), mesh(doog22), title('4');
% subplot(3,3,6), mesh(doog23), title('5');
% subplot(3,3,7), mesh(doog24), title('6');
% subplot(3,3,8), mesh(doog25), title('7');
% subplot(3,3,9), mesh(doog26), title('8');
% %
% figure,
% subplot(3,3,1), imscshow(dog2), title('1');
% subplot(3,3,2), imscshow(dog1), title('2');
% subplot(3,3,4), imscshow(doog21), title('3');
% subplot(3,3,5), imscshow(doog22), title('4');
% subplot(3,3,6), imscshow(doog23), title('5');
% subplot(3,3,7), imscshow(doog24), title('6');
% subplot(3,3,8), imscshow(doog25), title('7');
% subplot(3,3,9), imscshow(doog26), title('8');
% %
% imwrite(rescalegray(dog2), 'dog2.png');
% imwrite(rescalegray(dog1), 'dog1.png');
% imwrite(rescalegray(doog21), 'doog1.png');
% imwrite(rescalegray(doog22), 'doog2.png');
% imwrite(rescalegray(doog23), 'doog3.png');
% imwrite(rescalegray(doog24), 'doog4.png');
% imwrite(rescalegray(doog25), 'doog5.png');
% imwrite(rescalegray(doog26), 'doog6.png');
⛄ 运行结果
⛄ 参考文献
[1] Baseline JPEG and JPEG2000 Artifacts Illustrated, Aleks Jakulin, 2002, URI: http://ai.fri.uni-lj.si/~aleks/jpeg/artifacts.html
[2] S. Yang, Y. H. Hu, D. L. Tull, and T. Q. Nguyen, “Maximum likelihood parameter estimation for image ringing artifact removal,” IEEE Trans. Circuits and Systems for Video Technology, vol. 11, no. 8, August, 2001, pp. 963-973.
[3] A. Nosratinia, “Post-Processing of JPEG-2000 Images to Remove Compression Artifacts,” to appear in IEEE Signal Processing Letters.
[4] S. H. Oguz, T. Q. Nguyen, and Y. H. Hu, “Critical quantization decisions in transform coding and blocking artifacts,” Proc. ISCAS'99, Orlando, FL, pp. 63.19, 1999.
[5] S. Yang, Y. H. Hu, and D. L. Tull, “Blocking artifact removal using robust statistics and line process,” IEEE Int'l Workshop on Multimedia Signal Processing, 1999.
[6] S. Yang and Y. H. Hu, “Block Effect Removal using Regularization and Dithering,” Proc. ICIP'98, Chicago, IL, pp. 346-349, 1998.
[7] R. Krishnamurthy, J. W. Wood, and J. M. Francos, “Adaptive restoration of textured images with mixed spectra,” IEEE Trans. Image Processing, vol. 5, pp. 648-652, 1996.
[8] Y. H. Hu and R. Sambhare, “Constrained Texture Synthesis for Image Post Processing”, ICASSP '03, Hong Kong, 2003.
[9] J. Malik and P. Perona, “Preattentive texture discrimination mechanisms with early visio mechanisms”, J. Opt. Soc. Am. A. Vol 7, No. 5, May 1990
[10] R. Duda, P. Hart, and H. Stork, “Pattern Classification,” Wiley-Interscience, New York, NY, 2 ed., 2000.
⛄ Matlab代码关注
❤️部分理论引用网络文献，若有侵权联系博主删除
❤️ 关注我领取海量matlab电子书和数学建模资料
