dig_photo_2

特征提取中利用小波边缘检测C++源码，希望对大家有用，而且偶也能下载程序了，(feature extraction using wavelet edge detection C source code, we hope to useful, but can even download, Ha ha)

说明：  可以通过光流法来检测运动的目标，具有很好的效果，适合新手了解(It can detect moving target by optical flow method, and can detect moving target by optical flow method, which has good effect and is suitable for novice to understand)

说明：  图像处理，图像分割，图像变换，图像特征提取，图像增强，图像编码压缩(Image processing, image segmentation, image transformation, feature extraction)

背景建模和前景分割的方式把运动车辆提取出来。并进行最近临关联，输出目标轨迹。 MeanShift运动目标跟踪 matlab程序 1.截取跟踪目标矩阵rect 2.求取跟踪目标的加权直方图hist1 3.读取视频序列中的一帧， 先随机取一块与rect等大的矩形，计算加权直方图hist2。 4.计算两者比重函数，如果后者差距过大， 更新新的矩阵中心Y，进行迭代（MeanShift是一种变步长可以迅速接近概率密度峰值的方法），直至一定条件后停止。 这两种算法为处理图像跟踪的基本方法 (Background modeling and foreground segmentation the extracted moving vehicles. Pro associated output and the target locus. Target tracking matlab program MeanShift movement. Interception tracking the target matrix rect strike a tracking target weighted histogram hist1 3. Reads a video sequence, randomly take a with the rect large rectangular, calculate the weighted histogram hist2. 4. Calculate both the proportion of function, if the latter gap is too large to update the new center of the matrix Y conducted the iteration (MeanShift is a variable step size can be rapidly approaching the peak of the probability density) until certain conditions to stop. The basic method of these two algorithms for processing the image tracking)

一个阈值去噪程序，包括经典的去噪方法，如折中阈值，自适应阈值等等。(A threshold denoising procedures, including the classical denoising methods, such as compromise threshold, adaptive threshold and so on.)

说明：  各种分割算法的实验 包括阈值分割，多阈值分割，二维最大熵分割等(various experiments, including the threshold segmentation, multi-threshold segmentation, 2D Maximum Entropy segmentation)

凸一维总变差去噪与非凸正则化，内有文献参考。值得学习(One-dimensional convex total variation denoising non-convex regularization, the literature reference. Worth learning)

基于NSCT与HSI的图像融合算法，采用非下采样的Contourlet变换和HSI彩色图像融合。(image fusion algorithms based on NSCT and HSI)

训练了一个CNN模型用于土地资源分类，适用于遥感图像(A CNN model is trained for land use classification and is suitable for remote sensing images.)

CT图像的平行束重建图像的重建算法，非常高(CT image parallel beam reconstruction image reconstruction algorithm, very high)