双边去噪

Jsteg算法实现，Jsteg是一种基于JPEG的常用信息隐藏算法步骤如下: 首先，把掩体图像分为不重叠的8*8的子块，对每一子块进行DCT并对变换得到的DCT系数进行量化 其次，将待隐藏的信息进行加密，将加密结果嵌入到量化后值不为0，1或-1的DCT系数的最 低有效位(LSB)中，其嵌入顺序是按zigzag扫描顺序进行的。最后，用JPEG的嫡编码(包括哈夫曼编码，游程编码及DPCM)对嵌入秘密信息后的每一子块进行编码，从而得到一个含有秘密信息的JPEG stego文件。(Jsteg Algorithm)

在图像处理中实现带通滤波器的功能的matlab实现代码(Bandpass filter)

在指纹特征提取中，结合方向图和图像局部灰度方差对指纹图像进行分割，然后再进行阈值分割去处边缘效应得出较好的分割结果。(In the fingerprint feature extraction, the combination of pattern and image local variance of gray-scale fingerprint image segmentation, and then conducted threshold segmentation edge effects come to a better place for the segmentation results.)

利用卷积神经网络进行特征提取，并将卷积特征可视化(Feature extraction using convolutional neural networks and visualization of convolution features)

说明：  用FCM算法实现图像分割(图像源可以是灰度、索引和RGB图像)(FCM image segmentation algorithm (image source can be gray, indexes and RGB images))

HOG adaboost pedestrian-detect

利用互信息进行特征抽取的方法实现，能找到最具代表的特征，有效的实现了计算量的减少和可用来维数缩减(The use of mutual information feature extraction method of implementation, to find the most representative characteristics of an effective implementation of the computation of the reduction and can be used to reduce the dimension)

图像质量评价经典算法，ssim算法评价图像质量好坏程度的度量(Image quality assessment of the classical algorithm, SSIM algorithm.A measure of the degree of image quality )

说明：  这是一个用C++语言实现一个指纹识别程序，要求满足FVC国际竞赛标准，并在学校建立的指纹库上实测。程序里应用到的技术有：中值滤波、直方图均衡化、脊线方向提取、Gabor滤波、指纹细化、特征提取、特征点过滤、基于局部特征点的特征匹配(This is a C++ Language realization of a fingerprint identification procedure, the requirements FVC meet international competition standards, and established in the school library on the measured fingerprint. Program applied to the technology are as follows: median filter, histogram equalization, the direction of the ridge line extraction, Gabor filtering, fingerprint thinning, feature extraction, feature point filtering, based on local feature points of the feature matching)

实现了20000*20000以上的大tiff格式遥感图像的读写和显示，用于遥感图像处理(Achieve a 20,000* 20,000 or more large remote sensing image tiff format to read and write and display for remote sensing image processing)