codes_smc
所属分类：图形图像处理
开发工具：matlab
文件大小：5519KB
下载次数：29
上传日期：2013-08-29 17:36:53
上 传 者：
xingyu0312
说明： 基于支撑向量机的的图像质量评价方法，结合机器学习和图像特征提取
(The file feature_smc1.m is for feature extraction. It takes the reference and distorted images as input and gives the 256 dimensional feature vector as the output.
models_smc_revised.mat contains the 4 models developed by using TID, LIVE, CSIQ and watermark image databases.)
文件列表:
codes_smc_final\feature_smc1.m (2893, 2011-04-06)
codes_smc_final\models_smc_revised.mat (5819492, 2011-04-06)
codes_smc_final\svm\COPYRIGHT (1497, 2008-08-06)
codes_smc_final\svm\heart_scale.mat (28904, 2008-08-06)
codes_smc_final\svm\im (263222, 2009-05-19)
codes_smc_final\svm\kernel.dll (7680, 2006-01-05)
codes_smc_final\svm\kernel.m (1623, 2009-08-13)
codes_smc_final\svm\kernelproj.m (2112, 2009-08-13)
codes_smc_final\svm\kpca.m (3938, 2009-08-14)
codes_smc_final\svm\kpcarec.m (1568, 2009-08-13)
codes_smc_final\svm\make.m (208, 2008-08-06)
codes_smc_final\svm\Makefile (1326, 2008-08-06)
codes_smc_final\svm\pca.m (3408, 2009-02-02)
codes_smc_final\svm\ppdiff.m (1244, 2009-05-08)
codes_smc_final\svm\ppfit.m (10560, 2009-05-08)
codes_smc_final\svm\ppint.m (1662, 2009-05-08)
codes_smc_final\svm\read_sparse.c (2467, 2008-08-06)
codes_smc_final\svm\read_sparse.mexw32 (7680, 2008-08-06)
codes_smc_final\svm\speech (8184, 2008-08-07)
codes_smc_final\svm\splinefit.m (5627, 2009-05-15)
codes_smc_final\svm\svm.cpp (62021, 2008-08-06)
codes_smc_final\svm\svm.h (2899, 2008-08-06)
codes_smc_final\svm\svmpredict.c (8469, 2008-08-06)
codes_smc_final\svm\svmpredict.mexw32 (32768, 2008-08-06)
codes_smc_final\svm\svmtrain.c (10807, 2009-01-21)
codes_smc_final\svm\svmtrain.mexw32 (65536, 2008-08-06)
codes_smc_final\svm\svm_model_matlab.c (7632, 2008-08-06)
codes_smc_final\svm\svm_model_matlab.h (201, 2008-08-06)
codes_smc_final\svm (0, 2011-04-06)
codes_smc_final (0, 2011-04-06)
The file feature_smc1.m is for feature extraction. It takes the reference and distorted images as input and gives the 256 dimensional feature vector as the output.
models_smc_revised.mat contains the 4 models developed by using TID, LIVE, CSIQ and watermark image databases.
The folder "svm" consists of the support vector machine toolkit which has been downloaded from http://www.csie.ntu.edu.tw/\verb~cjlin/libsvm
Please follow the instructions for its installtion from their website.
********************************************************************************
An example usage is given below.
Suppose tr_l_a57 is the vector containing the subjective scores from the A57 database and tr_a57 consists of the corresponding image feature vectors.
load models_smc_final.mat
[pr_live,a1]=svmpredict(tr_l_a57,tr_a57',model_live_smc);
[pr_csiq,a1]=svmpredict(tr_l_a57,tr_a57',model_csiq_smc);
[pr_tid,a1]=svmpredict(tr_l_a57,tr_a57',model_tid_smc);
[pr_watermark,a1]=svmpredict(tr_l_a57,tr_a57',model_watermark_smc);
Then pr_live,pr_csiq,pr_tid and pr_watermark contain the predicted scores from the respective models.
*********************************************************************************
近期下载者：
相关文件：
收藏者：