pointevaluate

基于小波变换的图像融合源程序，通过调试，能直接使用。(Image fusion based on wavelet transform source code through the debugger can be used directly.)

说明：  一个目标检测识别的matlab程序，用于初学者使用matlab编程，用的比校全面(Identification of a target detection matlab program for beginners to use matlab to program than the school used a comprehensive)

用小波包分析实现轴承故障诊断，信号分析处理，界面编程(Bearing fault diagnosis using wavelet packet analysis and signal analysis and processing, interface programming)

针对于图像复制粘贴篡改检测，对图像数据进行降维处理，并进行块匹配，比较其图像块欧几里得距离(For copy and paste the image tamper detection, image data dimensionality reduction, and block matching, comparing its image block Euclidean distance)

通过区域迭代合并与定位切图的图像分割算法，非常好的。(Image Segmentation by Iterated Region Merging with Localized Graph Cuts)

全套的DCT域数字水印源代码matlab(A full set of digital watermarking of DCT-domain source code matlab)

图像重构实现干涉条纹缺陷图的缺陷检测，能很好的滤除条纹，得到缺陷(Interference fringe image reconstruction to achieve defect detection defect map can be a very good filter stripes)

说明：  本程序提出了变步长自适应滤波算法的步长调整原则:即在初始收敛阶段或未知系统参数发生变化时,步长应比较大,以便有较快的收 敛速度和对时变系统的跟踪速度 而在算法收敛后,不管主输入端干扰信号v ( n) 有多大,都应保持很小的调整步长以达到很小的稳态失调噪声. 根据变步长公式编的程序，很有参考价值. (This procedure, a variable step adaptive filter algorithm step adjustment principle: that in the initial convergence phase or unknown system parameters change, the steps should be relatively large in order to have fast convergence speed and time-varying systems tracking speed in convergence, no matter the main input interference signal v (n) how much should be adjusted to maintain a small step to achieve a small steady state misadjustment noise. compiled according to the formula variable step procedure useful reference.)

弹流油膜数值计算画等高图，难以得到预想马蹄形特征，通过代码处理，可以方便的解决问题，得到满意的预想图(EHL film numerical draw contour map, it is difficult to obtain desired characteristic horseshoe, processing through the code, you can easily solve the problem, get a satisfactory RENDERING)

采用了Harrtraining和离子滤波算法实现了拥挤人群人头计数，非常有用。VC++和Opencv开发，采用Oepncv中的Harrtraining训练一个分类器即可运行(Human detection and tracking in high density crowds is an unsolved problem. Standard preprocessing techniques such as background modeling fail when most of the scene is in motion. We integrate human detection and tracking into a single framework, and in- troduce a conrmation-by-classication method to estimate condence in a tracked trajectory, track humans through occlusions, and eliminate false positive detections.)