ImgHazeRemoval_Test

完成对视频帧的四叉树分割，在纹理复杂处分割块较小，在纹理平坦处分割较大(Complete quadtree segmentation of video frames, complex area in texture segmentation smaller, flat larger regional segmentation block in texture)

说明：  PCL+ MFC+ Visualization 点云显示(PCL+ MFC+ Visualization)

基于最小二乘法的相位解缠VC++算法 (least squares method based on the VC Phase Unwrapping Algorithms)

实现了三维超声PNN算法的GPU加速；实现GPU加速的CT圆轨道平板探测器FDK重建算法；提供对canny边缘检测算法的支持；提供交互式graph cut分割方法；实现OSTU分割算法；(Three dimensional ultrasound PNN algorithm GPU acceleration to achieve GPU accelerated CT reconstruction algorithm for circular orbits FDK flat panel detector provide support for canny edge detection algorithm provide interactive graph cut segmentation achieve OSTU segmentation )

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)

DOG检测算子，能够构造高斯尺度空间，并检测其中的极值，从而得到有价值的特征点位置(DOG detection operator, the Gaussian scale space can be constructed, and tested one of the extremes, resulting in valuable feature position)

自己毕业设计做的改进的基于某vc视频处理框架代码的重心多目标跟踪代码(The center of gravity of a vc video processing framework code-based target tracking code own graduation design improvements do)

人头检测 通过颜色分割产生人头头发部分的二值图像，计算此部分的面积和长宽比来判断是否有人头(human detection)

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)

对各种格式图像实现打开保存旋转的操作的示例工程的说明文档。(images of various formats to achieve preservation of open rotation examples of the operation of the project documentation.)