Image_Interpolation

sar图像变化检测 基于减法代码 比较两幅图片的细微差别(sar image change detection code based on subtraction of two pictures comparing the nuances of)

说明：  去噪模型TV,可用于图像去噪，效果良好，模型特别简单(The denoising model TV can be used for image denoising, the effect is good, and the model is particularly simple.)

说明：  机载合成孔径雷达自聚焦算法，没有实测数据，用的是点目标仿真(Airborne synthetic aperture radar algorithm for self-focusing, there is no measured data, using a point-target simulation)

C++，人脸变形，选择图片，在图片中选择一个四边形窗口（人物脸部），然后再在内部选择一个四边形窗口，逆时针选择，然后点击变形直接呈现变形结果，点击动画可以查看人脸变形过程(C++ program, select a picture, choose a quadrilateral window (human faces) in the picture, and then select a quadrilateral inside the window, counter-clockwise to select, then click on the deformation presented directly deformation results, click animated facial deformation process can be viewed)

说明：  利用大津算法对图片进行处理，得到满意的处理效果。(The image is processed by Otsu algorithm and satisfactory results are obtained.)

区域生长种子点的选取,可以用于图像分割的处理.(Simple single-seeded region growing)

利用传统的SVM-HOG算法，进行行人检测(HOG-SVM algorithm for pedestrian detection)

医学图像重建算法中OSEM算法的Matlab程序(MATLAB Program of OSEM Algorithms)

本程序是基于web图像的超分辨率重建，借鉴参考了其他人的算法(This procedure is based on the super image resolution of web image reconstruction, reference to other people s algorithm)

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.)