JPEG

这篇文章是一种特征降维的方法，不过跟一般的计较离散度的方法不同，这个特征变换的准则是：尽可能地保留原始样本的熵(Kernel Entropy Component Analysis Jenssen, R Pattern Analysis and Machine Intelligence, IEEE Transactions on Volume: PP , Issue: 99 )

基于小波和形态学的自然地面背景红外目标检测方法。。。。。。(Based on wavelet and morphology of the natural terrestrial background infrared target detection method)

图像处理的KMP算法源程序，KMP算法是一种基本的、实用的字符串比较算法，广泛应用于图像处理和模式识别。(Image processing algorithms source KMP, KMP algorithm is a basic, practical string comparison algorithm, widely used in image processing and pattern recognition.)

一个师兄编写的图像配准程序，其中sift代码匹配以及错误点检测值得参考学习。(A brother write image registration procedures, which sift code matches the reference point of error detection is worth learning.)

说明：  自己写的位图三态按钮还添加了按钮声音,对自己制作图形界面有很大帮助(Write your own bitmap three-state buttons to add the button sounds right to produce their own graphical interface helps a lot)

关于MC三维重建算法的二义性消除的文章，很好的消除了MC算法中的二义性(it s the article about Three-dimensional reconstruction algorithm to eliminate ambiguity in MC)

本程序可以使用图像处理算法对图片进行处理，对于一个含有各种物体的图片，如米粒，可以计算物体的个数。使用前请配置好opencv环境。希望对大家学习图像处理和opencv有所帮助...(This procedure can use the image processing algorithm for processing images for a picture with a variety of objects, such as a grain of rice, can calculate the number of objects. Configured before use opencv environment. We would like to learn opencv image processing and help ...)

ostu二值化算法vc实现，基于MIL图像处理软件包将图像解析，通过ostu求取二值化阈值，实现二值化算法(ostu binarization algorithm to achieve vc, MIL image processing software package based on the image resolution, strike through ostu binarization threshold binarization algorithm to achieve)

标记分水岭比较了开闭和基于重构的开闭的效果及分割染色后的效果图... 标记分水岭分割算法(Mark a watershed in comparing the opening and closing and effect diagram based on the effect of opening and closing and split reconstructed stained ... marked watershed segmentation algorithm)

一种简单高效地基于压缩感知的跟踪算法。首先利用符合压缩感知RIP条件的随机感知矩对多尺度图像特征进行降维，然后在降维后的特征上采用简单的朴素贝叶斯分类器进行分类。该跟踪算法非常简单，但是实验结果很鲁棒，速度大概能到达40帧/秒(A simple and efficient tracking algorithm based on compressed sensing. Firstly, with the random sensing matrix compressed sensing RIP conditions for multi-scale image feature dimension reduction, and then use the naive Bias classifier simple classification in the feature reduction after the. The tracking algorithm is very simple, but the results are robust, speed can reach 40 frames per second)