xindianxinhao

首先通过多分辨率分解法提取高频边缘，然后分别对每层的高频边缘进行非线性插值得到新的高频边缘，接着对高频边缘进行高通滤波得到等强后的高频边缘，最后对增强后的高频边缘进行修正。再在对增强的图像进行对比度拉伸，从而达到增强的目的！(First of all, through the extraction of multi-resolution decomposition of high-frequency edge, and then were on the edge of each high-frequency non-linear interpolation to be the new high-frequency edge, and then the edge of the high-frequency high-pass filter to be strong, such as after the high-frequency edge, and finally on the brink of enhanced high-frequency be amended. Again in contrast enhanced image stretching, so as to achieve the purpose of enhancing!)

说明：  IQA评估指标包括PLCC，SROCC，KROCC，PSNR，SSIM (Structural Similarity），MSE (Mean Square Error)，RMSE (Root Mean Square Error) ，信息熵 (Information Entropy)算法(IQA evaluation indicators include PLCC, SROCC, KROCC, PSNR, SSIM (Structural Similarity), MSE (Mean Square Error), RMSE (Root Mean Square Error) , Information Entropy (Information Entropy) algorithm)

数据库是由10个人的7种正面表情组成的213幅灰度图像，图像是以大小为256256的8位灰度级存储的，格式为.tiff型，平均每个人每种表情有2到4张。( the JAFFE database[11] which is composed of 213 images of female facial expression corresponding to 10 distinct subjects. Each image is stored at a resolution of 256×256 pixels and 8-bit gray level. Each subject in the database is represented with 7 categories of expression (angry, disgust, fear, neutral, sadness, happiness and surprise). )

图像检测的算法，Faster R-CNN算法，先对整张图像进行卷积计算，然后通过感兴趣区域池化层(RoI Pooling Layer)将选择性搜索算法推荐出来的候选区域和卷积网络计算出的特征映射图进行融合，得到候选区域对应的特征矢量，这种共享卷积计算的操作极大地减少了卷积计算的次数。而且这些特征矢量的维度统一，方便后续的分类工作。通过感兴趣区域池化层处理卷积特征，并将得到的特征送往两个并行计算任务进行训练，分类和定位回归。通过这些方法和改进的框架，Fast R-CNN 用更短的训练和测试时长，取得了比 R-CNN 更好的效果(Faster R-CNN algorithm first convolutes the whole image, then fuses the candidate regions recommended by the selective search algorithm and the feature mapping maps calculated by the convolution network through the RoI Pooling Layer to get the corresponding feature vectors of the candidate regions, which greatly reduces the number of convolution calculations. Moreover, the dimension of these feature vectors is unified, which facilitates the subsequent classification work. The convolution feature is processed by the pooling layer of the region of interest, and the obtained feature is sent to two parallel computing tasks for training, classification and positioning regression. Through these methods and improved framework, Fast R-CNN uses shorter training and testing time and achieves better results than R-CNN.)

混合像元分解，是定量遥感中非常重要的组成部分之一，在反演地表参数时是必须要考虑的因素之一，文中罗列了分解的一些方法。(mixed pixel decomposition, is the quantification of remote sensing very important component of one of the the inversion of surface parameters is to be considered one of the factors, the paper listed some of the decomposition method.)

局部二值模式，local binary pattern matlab代码(binary pattern matlab)

本文介绍了一种测量大气能见度的方法，双亮度差法，代码中给出了详细的注释，并且包括OSTU方法等等(In this paper, a measuring method of atmospheric visibility double brightness difference method, the code is given detailed annotations, and including the OSTU method and so on)

车牌识别。系统界面包括车牌图像预处理、定位、矫正、分割、识别等各个步骤的结果显示；显示包括待识别车辆和处理结果图像或数据(License plate recognition. System interface includes license plate image preprocessing, positioning, correction, segmentation, recognition and other steps of the results display; display includes the vehicle to be identified and processing results images or data)

对lagrange函数公式进行代码化，可以对一组数进行插值算法，可以输出最后的y值(The lagrange function formula code, a number of groups interpolation algorithm, you can output the final value of y)

说明：  SIFT用于图像全景拼接的源码，内附算法描述，很好用。(SIFT source code for image panorama stitching, containing algorithm descriptions, easy to use.)