搜索资源列表
二维小波变换
- 关于二维小波变换的程序 [精华] 说明:此算法重在概念,速度并不是很快。因为FOR循环的缘故。此程序从循环矩阵的观点出发,把圆周卷积和快速幅里叶变换建立了联系。实现了分解和无失真重构。它只做了一层分解,即将256x256图形分解成为64x64的四个图形,避免了使用WKEEP()的困惑。主要思想为用小波滤波器族构造正交阵W,变换写为B=W*A*W ,反变换为:A=W *A*W,这与所有正交变换无异。W为循环正交矩阵,因此可用FFT实现快速运算,难点就在重构矩阵上。若用矩阵概念明确,一个
Image_Classify
- 遥感影像分类的matlab实现(源码+图像)。以及分类后评价(总体精度、Kappa系数、混淆矩阵)。-Remote Sensing Image Classification of matlab implementation (source code+ images). After the evaluation and classification (overall accuracy, Kappa coefficient, confusion matrix).
ConfusionMatrix
- 遥感影像分类后精度评价C++程序,主要是通过混淆矩阵,包括生成者精度,用户精度。-Remote sensing image classification accuracy assessment after C++ program, mainly through the confusion matrix, including generators accuracy, user accuracy.
supervised_classification
- 模式识别中的监督分类程序,通过在遥感影像上选择训练样区,计算混淆矩阵,进行分类。-Pattern recognition supervised classification procedures, through training in remote sensing image, select the sample area, calculate the confusion matrix, classification.
jordixmarieeve.jpg
- confusion matrix for two actors performing 3 actions with 3 repetitions