Radon变换Matlab

进一步讨论怎么样进行虹膜眼睛的位置确认，区域生长法确定内圆轮廓(To further discuss how to confirm the position of the iris eye and determine the inner circle contour by region growth method)

已知一图像的轮廓，计算该轮廓质心，并标注在图上(A known the outline of image, the contour centroid calculation, and marked on the map )

各种信息熵的计算，包括Tsallis 熵，Renyi熵，Shannon熵等，及它们的扩展形式。(The functions include extensive Shannon and nonextensive Tsallis,escort Tsallis,and renyi entropy.)

sijiequzao 是台雪城老师的四阶去噪模型代码； subspace clustering accuracy of the cod 是子空间聚类正确率代码； TVdual 是ROF模型的原始对偶算法代码； TVH-1fj 是TV_H-1图像分解模型代码； TVl1fj 是TV_L1图像分解模型代码。(" Sijiequzao" is the teacher' s desk Syracuse Fourth Order denoising model code " subspace clustering accuracy of the cod" subspace clustering is the correct rate codes " TVdual" original ROF model dual algorithm code " TVH-1fj" is TV_H-1 image decomposition model code " TVl1fj" is TV_L1 image decomposition model code.)

MRF程序：建立了马尔科夫随机场 应用于图像去噪 不过此程序有待进一步调试(MRF procedure used in image denoising)

说明：  在MFC中显示各种格式的图片，jpg.png(Display pictures in various formats in MFC, jpg.png)

针对FCM算法的运行时间长和计算量大的问题，提出了改进的FCM算法，先将图像分割成窗口大小的子块，然后以子块为单位提取特征向量，用FCM聚类粗分割，然后对边缘子块，以像素为单位从新提取特征向量，进行细分割。分割后的结果提高了运行速度和分割精度。(For the FCM algorithm and the calculation of long run the problem of large proposed improved FCM algorithm, first image into blocks the size of the window, and then sub-block feature vector extraction unit, using FCM clustering coarse partition, and then block on the edge, in pixels from the new Feature Extraction, for fine segmentation. Improve the segmentation results after the speed and accuracy of segmentation.)

主动形状模型，实现轮廓提取有一定的帮助。(Active Shape Model)

L0梯度最小化算法，2011年由xuli提出的算法(L0 gradient minimization algorithm, in 2011 xuli put forward the algorithm )

针对可见光与红外图像的特点和难点，提出了可见光与红外图像配准与融合中的关键技术，即: 使用新型的基于一维最大类间方差和最大连通性测量的图像分割方法对源图像进行分割来更好地实行图像粗 配准 使用新型的特征点提取方法，特征点的匹配及误匹配的消除来更好地实行图像精配准 采用新型的基 于区域的树状小波活性测度计算来实现树状小波图像融合 利用自生成神经网络来实现模栩图像融合. (For visible light and infrared images of the characteristics and difficulties, a visual and infrared image registration and integration of key technologies, namely: the use of new one-dimensional based on the largest variance between-class connectivity and the largest measuring method of image segmentation source image segmentation to better implementation of coarse image registration the use of a new type of feature point extraction method, feature points of the match and the elimination of false matches to better the implementation of image fine registration the use of new tree-based wavelet calculated to measure the achievement of the activity tree wavelet image fusion the use of self-generating neural network to achieve image fusion Xu mode.)