pcnn

利用HOG算子提取特征，利用支持向量机进行分类，得到了较好的图像分割效果(Using HOG operator to extract features, using support vector machine to classify, get a better image segmentation effect)

基于matlab的、结合深度图像的图像修补( Image patch based on matlab, combined with the depth of the image )

用matlab写的人脸检测代码,具有很好的识别率，对做人脸识别，检测的研究有一定的帮助(the code is used to face detaction and it has a high rate.It is very useful for the search of face detaction )

将原图像进行退化处理（选择大气湍流导致图像模糊时相应的PSF），然后将退化后的图像进行复原（使用维纳滤波、直接逆滤波）.(The original image degradation process,and then restore the degraded image with Wiener filtering and inverse filtering.)

从回波中估计多普勒参数是合成孔径雷达运动补偿的关键。文中研究了多目标回波对参数估计的影响，从理论上推导了估计多普勒中心和调频率对回波的不同要求，并用两个指标举例分析了一段实测机载SAR数据，给出了实际应用中参数估计时选择回波的原则。(Doppler parameters estimation from radar echoes is a key step in SAR(Synthetic Aperture Radar}motion cornpensation． This paper researched the efect of multiple targets echoes to estimate parameters and deducted theoretically the different requirement for echoes in estimating doppler centroid and doppler frequency modulation rate．Th en giving a real airborne SAR echo as example，we analyzed it with two measurements and concluded the principle about choosing echoes in practice．)

此程为目标检测算法程序，采用经典的光流算法，最后可以的到一个运动光流图。(This process procedures for the target detection algorithm, using the classical optical flow algorithm, and finally to a movement to the optical flow diagram.)

说明：  MTLAB编写的程序，包括形态学梯度检测二值图像的边缘、用水线阈值法分割图像等。(MTLAB preparation procedures, including detection of morphological gradient binary image of the edge of the water line thresholding segmentation images.)

实现了一个基于颜色、纹理、形状综合特征的MatlabGUI图像检索系统。(Bulit an image retrieval system based on which combines these three features and realizes user-defined weights.)

替代数据法中最重要的文献和文献相关的软件代码,非常经典亚(Surrogate data method is the most important thing in literature and literature related software code, very classic)

图像分解算法，四种方法可选，包括threshold, K means threshold, region growing以及gradient based segmentation(Image decomposition algorithm, four methods available, including the threshold, K means threshold, region growing and gradient based segmentation)