Hydrid_GA_ACATSP

利用TMS320VC5509A为核心的DSP处理硬件,通过软件实现自适应Notch滤波和瞬时频率序列方差检测(VIRD)运算,在较低信噪比使用条件下完成对单频测距信号的可靠检测和判定(Using tms320vc5509a DSP core processing hardware, adaptive notch filter and variance of instantaneous frequency sequence detection (Vird) operation is realized by software, in low signal to noise ratio under the conditions of use complete reliable detection of single frequency ranging signal and judging)

说明：  该压缩包内含有基于梯度金字塔变换的多模态图像融合方法的MATLAB代码，可以直接运行，得到融合结果。另外，压缩包内包括了多组已配准、待融合的图像示例，具体包括红外图像与可见光图像、CT与MR图像，多聚焦图像。本人还对代码进行了必要的解释。推荐下载(The compression package contains matlab code of multimodal image fusion method based on gradient pyramid transformation, which can be run directly to get fusion results. In addition, the compressed package includes several groups of images that have been registered and to be fused, including infrared images and visible images, CT and MR images, and multi focus images. I also gave the necessary explanation to the code. Recommended download.)

如题，里面有414张人脸图+matlab 图像重构代码和图像主成分代码(As the title says, there are 414 face pictures in it.+Model of image rebuild +Pca for Image code)

说明：  此程序为地震层析成像的LSQR算法，此算法用来解反问题。(This procedure for seismic tomography LSQR algorithm, the algorithm used to solve inverse problems.)

鼠标滑动截图，截图后 点击winform其它位置 则将图片黏贴到相应位置

自编语谱图函数，可实现基本的spectrogram功能(Self compiled spectral function, which can implement basic spectrogram functions)

这些代码主要是IDL语言源码，主要功能：初始化ENVI,定标,大气校正,融合,镶嵌,裁剪,分类(监督+非监督),面向对象特征提取,结束ENVI(These codes are mainly IDL language source code, the main functions: initialization, ENVI, calibration, atmospheric correction, fusion, mosaic, cutting, classification (supervision + non supervision), object oriented feature extraction, ending ENVI)

改进的sift算法，称为surf，来实现图像的匹配，并显示。 (Improved sift algorithm, known as the surf, to achieve the image matching, and displayed.)

基于Curvelet变换的样本块图像修复算法提高现有样本块修复算法性能。首先利用Curvelet 变换估计待修复图像的4方向特征．然后利用颜色信息与方向信息共同衡量样本块间的相似度，在此基础上构造颜色-方向结构稀疏度函数．(Based on the sample block transform image restoration algorithm Curvelet enhance existing sample block repair algorithm performance. Firstly Curvelet transform estimated to be repaired 4 directional characteristics of the image. Then use the color information and direction information together measure the similarity between the sample block, constructed on the basis of color- direction sparsity structure function.)

用于火焰尖角、圆形度、纹理的测量以及图像颜色、矩阵、边缘直方图的测量(For flame sharp corners, roundness, texture measurements, and image color, matrix, edge histogram measurements)