搜索资源列表
USPS.rar
- 用于手写体数字识别的USPS样本数据库和将MAT格式的样本数据库转换成二值化图像并以行程编码存储,For handwritten numeral recognition of the USPS database and a sample format of samples MAT database into binary image and the Run-Length Coding storage
Ncut_SVM
- 此源码可对图像进行Ncut分割，并且集成了特征提取和SVM分类的功能。-This source can be Ncut image segmentation, and integrates the feature extraction and SVM classification function.
m-
- 用小波变换进行图像预处理，再用自编小程序寻优参数，并用svm分类-Using wavelet transform for image preprocessing, and then writing small programs optimization parameters, and SVM classification
Seal-Identification
- 运用H15；I色彩窄间对印章图像颜色特征提取等一系列预处理，研究了多种识 别方法后发现基于纹理特征的印鉴识别方法不但速度快，而且识别率也较高，并且方法简便。通 过将极坐标和傅垦叶变换结合，计算m印签图像纹理的频谱度量，来完成印鉴的特征提取，运用支 持向量机分类器对印鉴进行识别，实验证明，方法具有良好的旋转不变纹理分析性能，提高了识别 率。-It WaS found a fast，simple and a higher seal image(HSI)identification
Linear-classifier-design
- 对“data1.m”数据，分别采用感知机算法、最小平方误差算法、线性SVM算法设计分类器，分别画出决策面，并比较性能。-The "data1.m" data, respectively, using the perceptron algorithm, the least square error algorithm, the linear SVM algorithm design classifier, respectively, to draw the decision-making surf
voc-release3.1(gai)
- 这个就是SVM识别的，说是世界识别冠军程序。源程序不是在windows 下的，这个该成了Windows下的，打开DEMO.M文件直接运行。检测的程序改好了，-This is the SVM identified, said that the world championship recognition program. The source is not in the windows, which became the Windows, open the file directly DEMO.M
BRISQUE_release
- This a demonstration of the Blind/Referenceless Image Spatial Quality Evaluator (BRISQUE) index. The algorithm is described in: A. Mittal, A. K. Moorthy and A. C. Bovik, No Reference Image Quality Assessment in the Spatial Domain Binaries:
code-faceGUI.m
- 图形界面的人脸识别代码，PCA+SVM多分类，yale图片库-Face recognition codes in GUI, PCA+SVM multi-classification, yale gallery
基于支持向量机的图像分类
- classifier.mat文件太大未上传。可运行一次main3生成 getFeatures.m 获取灰度共生矩阵相关特征 main-main6 训练 + 识别 predict.m 单独的分类程序 temp.m 单独的分类演示程序(classifier.mat file is too large to upload. Can run a main3 to generate getFeatures.m Get the grayscale co-occurrence matrix rel