灰度图像边缘检测

批量读取tiff影像并计算月份NDVI最大值(use matlab to read tiff and calculate NDVI)

该程序主要功能是实现两幅图像特征点的匹配。首先用harris角点检测算法检测出图像角点，然后使用ssd算法筛选匹配的角点，最后将匹配的角点用线连接起来(The program is to achieve two main functions of image feature points of the match. First, with harris corner detection algorithm to detect image corners, then use the ssd corner matching algorithm selection, the final match of the corner will be connected with a line)

基于高斯肤色模型检测肤色，可用于人脸检测(based gaussian skin model detect skin can be used face detect)

先检测出椭圆，然后对椭圆进行椭圆拟合，画椭圆。(First detected in an oval, and then carried out on the elliptical ellipse fitting, painted oval.)

图像数字水印的方案毕业论文，基于dct和dwt的算法，有详细程序。(digital watermark image of the program thesis, based on Extra dwt and the algorithm, a detailed procedure.)

信息检索和自然语言处理中经常会使用这些参数:准确率（Precision）、召回率（Recall）以及综合评价指标（F1-Measure ） (These parameters are often used in information retri and natural language processing (Precision), recall rate (Recall), and comprehensive uation index (F1-Measure). )

　　本函数选用阈值分割的方法来完成瞳孔定位。根据图像的灰度分布特性，瞳孔部分的灰度值最小，其次为虹膜部分，最大的为巩膜部分。因此在灰度直方图中，第一个大峰值的邻域即是瞳孔所在区域。(Function selects the threshold segmentation method to accomplish the pupil location. Based on image gray distribution features, the pupil part of the grey value of the minimum, followed by the iris part, the biggest part of sclera. So in gray level histogram, the peak value of the first big neighborhood is the pupil area.)

Harris角点检测，关于角点的应用在图像处理上比较广泛，如图像匹配(FPM特征点匹配)、相机标定等。网上也有很多博客对Harris角点检测原理进行描述，但基本上只是描述了算法流程，而其中相关细节并未作出解释，这里我想对有些地方做出补充说明，正所谓知其然知其所以然，如有不对，还望指正。(Harris corner detection)

视觉slam十四讲，程序资料：g2o优化、cere拟合、特征点提取与匹配、点云拼接、opencv图像读取(Visual slam 14, program data: g2o optimization, cere fitting, feature point extraction and matching, point cloud mosaic, OpenCV image reading)

说明：  改进LBP纹理提取以及纹理分类，主要包括三部分测试：纹理提取、分类和分割(Improved LBP texture extraction and texture classification, including three parts of testing: texture extraction, classification and segmentation)