mine

shi-tomasi方法角点提取方法。该方法测量结构张量的最小特征值。这代表了Harris和诺贝尔探测器试图近似的理想。哈里斯希望在计算特征时避免计算平方根的计算成本。(Shi-Tomasi corner detector method.The method measure returns the minimum eigenvalue of the structure tensor. This represents the ideal that the Harris and Noble detectors attempt to approximate. Back in 1988 Harris wanted to avoid the computational cost of taking a square root when computing features. This is no longer relevant today!)

运用HOG提取人脸特征与支持向量机算法训练模型，进行人脸识别(Using HOG to extract facial features and support vector machine algorithm training model for face recognition)

说明：  用于绘制节理面粗糙度风玫瑰图的Matlab程序，包含节理面粗糙度各向异性的案例数据！(Matlab program for drawing wind rose diagram of joint surface roughness, including case data of joint surface roughness anisotropy)

摄象机标定的两步法修改程序代码，标定摄象机内外参数(Two-step camera calibration procedure to amend the code, both within and outside the parameters of camera calibration)

实现了车辆目标的背景建模、前景检测、阴影消除、车辆跟踪(Achieve a vehicle target background modeling, foreground detection, shadow elimination, vehicle tracking)

对图像进行GMM（混合高斯）拟合后用EM算法进行分割(Image GMM (Gaussian Mixture) after fitting algorithm using EM Segmentation)

计算一幅灰度图像的直方图；进行简单的线性变换，并与源图像的直方图做变换；利用直方图均衡化进行图像增强(Calculating a gray-scale image histogram simple linear transformation, and to do with the source image histogram transformation the use of histogram equalization for image enhancement)

这是一个地震数据多尺度分析和处理的文章.内容相当丰富,不仅仅增加地质方面的知识,更重要的是可以学到很多数据处理和小波分析方面的知识.(This is a multi-scale seismic data analysis and processing of the article. Fairly rich, is not only increasing geological knowledge, the more important it is to learn a lot of data processing and wavelet analysis knowledge.)

提出了一种基于非下采样Contourlet 变换的自适应图像去噪方法 首先对噪声图像进行非下采样Contourlet 变换，得到各 个尺度各个方向子带的系数，再根据该系数的能量自适应地调整去噪阈值 实验表明，与Contourlet 多尺度阈值去噪 Contourlet 自 适应阈值去噪相比，该方法在保留图像边缘细节的同时，能提高图像的PSNR 值，减少了Gibbs 现象(A new adaptive method of image de-noising based on NonSubsampled Contourlet Transform （NSCT）is presented.Firstly， the noisy image is decomposed into a set of multiscale and multidirectional frequency subbands by NSCT， and then the threshold is adapted automatically according to the energy of subband coefficients.Compared with the multi-scale threshold using contourlet transform and using adaptive threshold based on contourlet transform， the simulation results show that the performance of this method is superior in PSNR， meanwhile it can maintain the image edge and reduce the Gibbs phenomen)

说明：  图像去噪，降低图像的噪声，提高图像PSNR(image denoise,improve the psnr of image)