Compressive-Classifier

利用数字图像处理软件halocn来进行缺陷检测，附带有之前做工程项目的原图源代码，还有ppt和实验指导书(Halocn using digital image processing software for defect detection, comes with artwork done before the project source code, as well as ppt and experimental instructions)

非线性CS算法,提高SAR成像的精度，弥补CSA和RDA的不足(Nonlinear Chirp Scaling Algorithm)

说明：  在 matlab中实现一种局部阈值化图像的算法(In matlab to achieve a local thresholding algorithm for image)

说明：  hdr图片合成，四幅不同曝光的图片合成一张hdr图片(HDR picture synthesis, four pictures of different exposures synthesize an HDR picture)

包括基于SIFTGPU的sift特征提取，sift匹配，基于ransac误匹配点剔除等(SIFTGPU based sift feature extraction, sift matching, mismatching points based ransac removed, etc.)

用matlab实现人眼定位的程序 ，包括滤波、阈值分割、人眼粗定位、精确定位等。(Using matlab realize the human eye positioning procedures, including filtering, threshold segmentation, the human eye coarse positioning, precise positioning.)

针对空间发光点缩成光斑图像中心的高精度提取问题，本文提出一种亚像素提取方法。该方法简历了具有高斯能量分布的空间发光点的透视投影光斑图像灰度分部数学模型，证明了空间发光点能量中心的透视投影不变性。(Light-emitting points of space shrunk to spot high-precision center of the image extraction problem, we propose a subpixel extraction method. This method resumes a space with Gaussian energy distribution points of perspective projection spot light gray segment mathematical model to prove the energy center of the space light-emitting point perspective projection invariance.)

Canny边缘检测算法,canny算法计算梯度,并从小波模极大重构图像.(Canny edge detection algorithm, canny algorithm for calculating the gradient, and from the Wavelet Modulus Maxima Image Reconstruction.)

该代码实现序列图像配准和重建过程，包含多种算法，配准算法有keren,marcal等，重建算法包含了POCS凸集投影算法(The code sequence of image registration and renewal process, including a variety of algorithms, registration algorithms are keren, marcal etc. reconstruction algorithm contains POCS POCS algorithm)

该程序主要利用快速傅里叶算法将振动信号的时域转化成频域信息，得出频谱进行分析。(The program mainly using fast Fourier algorithm time domain vibration signal converted into the frequency domain information, obtain spectral analysis.)