hanfeng

通过边缘检测寻找图像上圆点的圆心坐标，可用于大量的图像处理程序(Image through edge detection to find the center coordinates of the dots)

基于分块的人脸图像超分辨率学习算法实验平台，利用学习到的训练集图像信息对低分辨率图像超分辨率，并可以查看超分辨率后的图像块匹配结果（用鼠标双击结果显示窗口中的实验图像），可以设置实验参数。(Block-based face image super-resolution algorithm for the experimental platform to learn, learning to use the training set of low-resolution image information super-resolution images and can view the image super-resolution block-matching results (double-click with the mouse results display window of the experimental images), the experimental parameters can be set.)

说明：  全变分去噪程序，用于图像去噪，该方法能克服图像去噪中的震荡效应，使图像更加光滑，并能保存边缘信息。(total variation denoisy)

人脸图像配准，将同一场景的两幅或多幅图像进行对准，可以实现人脸归一化(Face image registration, the same scene two or more images for the alignment can be achieved face normalization)

对于图像不同纹理特征值（能量、熵、惯性矩、相关）的提取(For different image texture feature value (energy, entropy, moment of inertia, related) extraction)

本程序主要实现低分辨率图像的超分辨率重建，有助于我们理解图像重建的过程和原理(This procedure is mainly to achieve super-resolution reconstruction of low-resolution images help us understand the process and principles of image reconstruction)

说明：  在图像恢复技术中，总变分最小化是一种以保存图像细节为目标的规整化复原方法。本算法基于总变分最小化的思想，采取交替最小化的精确迭代策略来对图像和点扩展函数同时进行恢复。实验表明，该算法在噪声水平较高的情况下仍能对图像进行稳健的恢复。(Total variation image restoration techniques is a goal to save the image detail regularization recovery method. Algorithm based on total variation minimization thinking, take the alternating minimization accurate iteration strategy to simultaneously restore the image and point spread function. The experiments show that the algorithm in the case of high noise levels can still robust recovery image.)

李纯明博士的无需初始化模型的源代码，作者亲自写的。还有测试图象，相信对大家肯定有好处。(Economics Dr. initialization model without the source code, written by the author himself. There test image, I think we will benefit.)

说明：  使用gpu、cpu并行进行sift算子计算匹配，能够在原来的基础上加速处理，但对显卡要求较高，具体环境配置使用方法可以参照mannual(SiftGPU is an implementation of SIFT [1] for GPU. SiftGPU processes pixels parallely to build Gaussian pyramids and detect DoG Keypoints. Based on GPU list generation[3], SiftGPU then uses a GPU/CPU mixed method to efficiently build compact keypoint lists. Finally keypoints are processed parallely to get their orientations and descriptors. SiftGPU is inspired by Andrea Vedaldi s sift++[2] and Sudipta N Sinha et al s GPU-SIFT[4] . Many parameters of sift++ ( for example, number of octaves, number of DOG levels, edge threshold, etc) are also available in SiftGPU. The shader programs are dynamically generated according to the parameters that user specified. SiftGPU also includes a GPU exhaustive/guided sift matcher SiftMatchGPU. It basically multiplies the descriptor matrix on GPU and find closest feature matches on GPU. Both GLSL and CUDA implementations are provided. )

随着红外成像技术的发展，红外小目标检测在军事领域有越来越重要的应用。以此为背景，研究了天空、海面、大地等各种背景下红外视频中弱小目标的检测问题。 (With the development of infrared technology,dim small targets detection play a more and more important part in the military. The problem of dim small targets detecting in a infrared video with the background of the sky,the sea and the groundhas been studied. )