cluster

对图像添加各种噪声 高斯噪声 淑盐噪声(Add a variety of image noise, Gaussian noise, noise, salt-sook)

说明：  通过IDL，简单实现灰度变换，彩色变换等基本操作，包括图片打开保存拖拽显示(Simple implementation of gray level transformation, color transformation and other basic operations)

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.)

运用openCV编写的C++图像色彩转移程序，原理是先转化到Lab图像，再保持L不变调整图像的a、b分量的均值跟方差。(Using openCV prepared C++ image color transfer process, the principle is first converted to Lab image, and then adjust the image to keep L constant a, b, with the mean variance component.)

求解图像的穆勒矩阵及其极分解矩阵，还有矩阵分析参数(Muller matrix and polar factorization matrix for solving image)

快速k-means算法,比matlab 自带的要快很多,不错的源码(K- means algorithm quickly, much faster than with matlab, good source)

基于条件迭代法与马尔可夫随机场的遥感图像分割的matlab代码(the image segmentation based on the ICM and Markov random field)

提出了一种基于图像边缘连接的线段检测方法。数字图像中的线段是由很多线段图元组成,在同一 条线段中,它们的方向是相同的并且受到真实线段的方向约束。因此可以通过提取小的线段图元, 然后分析它们 的连接特性, 进行线段的检测。首先用ADM 算法对图像进行边缘检测, 得到单像素宽的边缘 然后在4 个方向上 进行线段的检测 对初步得到的结果进行合并和验证, 去掉短的线段, 连接临近的线段。经过实验证明, 该方法能 够快速地提取图像中的线段,并且对噪声具有一定的鲁棒性(A line segment connecting a detection method based on image edges. The segment of the digital image is composed by many segments elements, their orientation is the same and by the real line segment direction constraint in the same segment. By extracting a small segment Element can then analyze their connection characteristics for the detection of the line segment. First with ADM algorithm for edge detection on the image to obtain a single-pixel-wide edge then in the four directions on the detection of the line segments merger and verification of the preliminary results obtained, removing the short line segments, the line segments connected near. The experiments show that this method can quickly extract the line segments in the image, and certain robustness to noise)

基于形态学的权重自适应图像去噪；通过串并联的滤波降噪对比图，说明并联降噪的优越性(Heavy adaptive image denoising based on morphology right noise filtering by series-parallel comparison chart, show the advantages of parallel noise reduction)

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). )