LT1

Image Texture Classification Using Combined Grey Level Co-Occurrence Probabilities and Support Vector Machines Texture refers to properties that represent the surface or structure of an object and is defined as something consisting of mutually related elements. The main focus in this study is to do texture segmentation and classification for texture digital images. Grey level co-occurrence probabilities (GLCP) method is being used to extract features from texture image. Gaussian support vector machines (GSVM) have been proposed to do classification on the extracted features. A popular Brodatz texture album had been chosen to test out the result. In this study, a combined GLCP-GSVM shows an improvement over GLCP in terms of classification accuracy.

说明：  遗传算法，基于二进制编码形式，用于解决优化问题，具有很好的全局优化效果。(genetic algorithm for optimization)

my final thesis on work of DSK development i included all the source code here.

Given gray scale double matrix, M(width,height,frameNum),this function stabilizes the sequence using the Grey-Coded Bit-Plane Matching (GC-BPM) algorithm.(Given gray scale double matrix, M (width, height, frameNum). this function stabilizes the sequence using th Gray e-Coded Bit-Plane Matching (GC-BPM) algo rithm.)

Useful Paper for my research work

说明：  信号处理,进行信号检测与处理，通过设定threshold来进行信号检测的可信度调整(Credibility adjustment signal processing, signal detection and processing, by setting threshold for signal detection)

是自适应阈值的分割，主要对图像用matlab进行分割图像，确定图像想要的东西(Adaptive segmentation threshold, the main image to image segmentation using matlab, determine what image you want)

形态学中的腐蚀和膨胀运算。首先，对图像进行低通滤波，消除细小噪声的干扰。 在滤波时，没有采用传统的中值滤波器，而是采用顺序统计滤波器ordfilt2(Erosion and dilation operation in morphology. Firstly, the image is filtered by low pass filtering to eliminate the interference of small noise. In filtering, the traditional median filter is not used, but the order statistics filter ordfilt2)

用0－1test方法检测混沌的程序，程序完整，运行良好。(0－1Test)

本程序利用奇异值分解对3通道彩色图像进行压缩分解，具体步骤如下： 压缩过程： 1. 选取子图像大小K值，把图像分解成M×M个子图像，IMG(s),s=1,2,…, M2，其中M=N/K，原始图像IMG大小为N×N。 2. 计算这M2个子图像的平均值average，对每幅子图像减去均值图像得到新图像。 3. 计算相关矩阵R，其元素定义为 。 4. 计算R的特征值与特征向量，计算每幅子图像与最大特征向量的内积，便得到编码，即压缩后的图像。 (This procedure using singular value decomposition of 3-channel color image compression decomposition, concrete steps are as follows: compression process: 1. Select the sub-image size K value, the image is decomposed into M × M sub-image, IMG (s), s = 1, 2, ..., M2, in which M = N/K, the original image IMG size N × N. 2. M2 calculate the average of sub-image average, for each sub-image minus the new images mean images. 3. Calculation of correlation matrix R, defined as its elements. 4. Calculation of the characteristics of R value and eigenvector calculated each sub-image with the largest eigenvector of the inner product, they will have to code, that is, the compressed images.)