codebar

求解SVM问题的一种高效算法,每次更新两个a()

基于Fisher线性判别基础学习器的集成分类器专门用于数字媒体的隐写分析，目前使用高维特征空间。目前，由于其良好的检测精度和较小的计算成本，它可能是设计监督分类器用于数字图像隐写分析的最常用方法。(The ensemble classifier, based on Fisher Linear Discriminant base learners, was introduced specifically for steganalysis of digital media, which currently uses high-dimensional feature spaces. Presently it is probably the most used method to design supervised classifier for steganalysis of digital images because of its good detection accuracy and small computational cost. It has been assumed by the community that the classifier implements a non-linear boundary through pooling binary decision of individual classifiers within the ensemble.)

多尺度多方向性的剪切波Shearlet变换优于小波变换，里面包含很多例子，可以直接运行，可以用于图像去噪，分割，增强等，是一个优于小波的图像稀疏表示工具(Shearlet transform of multi-scale and multi-directional shear wave is better than wavelet transform. It contains many examples and can run directly. It can be used for image denoising, segmentation and enhancement. It is an image sparse representation tool better than wavelet transform.)

鱼眼图像校正并将校正后的图像转换成鸟瞰图，即俯视图，纯C代码，没有调用其它库(fish eye image correction,then change it to bird s eye view)

利用GS算法实现相位的校正，从而使相位得到一定程度的改善(The phase correction is realized by GS algorithm, so that the phase is improved to a certain extent.)

对移动目标检测，需要先用鼠标框定感兴趣的目标，然后才能追踪。(Moving target detection, target mouse framed interested need to use before tracking.)

通过在matlab软件平台上进行仿真，实现天线阵列的自适应波束赋形(Realization of adaptive beamforming)

说明：  logistic混沌映射图，可以得到对logistic混沌映射图和基于logistic对图像加密(Logistic chaotic map)

说明：  lowe论文的sift算法的MATLAB源代码(MATLAB source code of sift algorithm for lowe paper)

针对目前工业制造领域面临的难题,提出利用非量测数字摄像机进行工业钣金件高精度三维重建与视觉检测.采用二维直接线性变换分解摄像机参数初值并结合光束法平差进行高精度标定 利用最小二乘模板匹配提取像面上的点、直线信息并进行混合光束法平差,从而进行钣金件的高精度三维重建及其尺寸制造误差的检测.所开发的视觉检测系统硬件成本低、自动化程度较高,实际数据的多次实验均取得了优于1/8 000的相对精度,说明所论述的方法为工业零件的自动化三维检测提供了一条可行途径.(View of the current challenges facing the field of industrial manufacturing, made use of non-metric digital camera precision sheet metal parts for industrial reconstruction and visual inspection. Using two-dimensional direct linear transformation and the decomposition of initial parameters of the camera bundle adjustment with high accuracy calibration extracted by the least square template matching, like the surface of point, line information and mixed bundle adjustment to the three-dimensional reconstruction of high-precision sheet metal parts and the size of the detection of manufacturing errors. visual inspection system developed by the hardware low cost, high degree of automation, the actual data, many experiments have achieved better than 1/8000 of the relative accuracy, that the methods discussed in the automation of industrial components to provide a feasible three-dimensional detection approach.)