Glcm-untitled

this is a project about animation. how do we form the shape to make it look like car, house abd so on. with the colours that wee like and the movement.

该程序初始时使用背景差分法自动检测目标，当目标完全进入图像中后，改用meanshift方法对目标进行跟踪。(the program use background subtraction to detect the target first,when the target fully enter the screen,change to meanshift method to track the target.)

实现原始的JPEG图片编解码，不借助任何的第三方库，纯C++。(To achieve the original JPEG image decoding, without using any third-party libraries, pure C++.)

水平集分割图像的一种算法，属于比较前沿的一种图像分割方法。(A level set image segmentation algorithm, a relatively cutting edge of a segmentation method.)

matlab 归一化相关系数，可用于模板匹配(matlab normalized correlation coefficient)

说明：  对第一个源码修改后的9月10日的Matalb的基于Adaboost算法的车牌定位的源码！ 里面含有积分图像的获得，haar 特征的生成等。感兴趣的欢迎下载！(Source of the first amended September 10 of Matalb-based license plate location algorithm Adaboost source! Which contains points of access to images, haar generation characteristics. Interested welcome to download!)

本文以自主驾驶车辆为实际应用背景．提出了一种改进的ＲＲＴ（快速随机搜索树）路径规划算法．该路 径规划算法将非完整性约束条件与双向多步扩展ＲＲＴ搜索算法相结合，在提高搜索效率的同时保证了规划路径的可 行性．同时将路径点作为Ｂ样条基函数的控制点，用三次Ｂ样条函数来拟合控制点生成平滑可跟踪的路径．通过在平 面障碍物环境下实验，验证了该算法的有效性．(ＡｉｍｉｎｇａｔｔｈｅａｐｐｌｉｃａｔｉｏｎｏｆＡＬＶ（ａｕｔｏｍａｔｉｃｌａｎｄｖｅｈｉｃｌｅ），ａｎｉｍｐｒｏｖｅｄＲＲＴ（ＲａｐｉｄｌｙＲａｎｄｏｍ?ｅｘｐｌｏｒｉｎｇ Ｔｒｅｅｓ）ｐａｔｈｐｌａｎｎｉｎｇａｌｇｏｒｉｔｈｍ ｉｓｐｒｏｐｏｓｅｄ．Ｔｈｅａｌｇｏｒｉｔｈｍ ｃｏｍｂｉｎｅｓｔｈｅｎｏｈｏｌｏｎｏｍｉｃｃｏｎｓｔｒａｉｎｔｓｏｆｖｅｈｉｃｌｅｗｉｔｈｄｏｕｂｌｅｅｘｔｅｎｄ ＲＲＴｓ，ｗｈｉｃｈｎｏｔｏｎｌｙｉｍｐｒｏｖｅｓｔｈｅｅｆｆｉｃｉｅｎｃｙｏｆｓｅａｒｃｈｉｎｇｂｕｔａｌｓｏｇｕａｒａｎｔｅｅｓｔｈｅｆｅａｓｉｂｉｌｉｔｙｏｆｔｈｅｐａｔｈａｔｔｈｅｓａｍｅｔｉｍｅ．Ｔｈｅｐａｔｈ ｐｏｉｎｔｓａｒｅｔａｋｅｎａｓｒｅｆｅｒｅｎｃｅｐｏｉｎｔｓｆｏｒｔｈｅＢ?ｓｐｌｉｎｅｂａｓｉｃｆｕｎｃｔｉｏｎ，ａｐｐｒｏａｃｈａｎｄｃｒｅａｔｅａｎｅｗｓｍｏｏｔｈｒｏｕｔｅｗｈｉｃｈｓｕｉｔｓｆｏｒｔｈｅｖｅｈｉ? ｃｌｅｔｏｆｏｌｌｏｗ．Ｔｈｅｓｉｍｕｌａｔｉｏｎｏｕｔｃｏｍｅｓａｎｄｆｉｅｌｄｔｅｓｔｓｖｅｒｉｆｙｔｈｅｅｆｆｅｃｔｉｖｅｎｅｓｓｏｆｔｈｅｍｅｔｈｏｄｐｒｏｐｏｓｅｄ．)

说明：  利用其中不同的光谱匹配算法实现光谱匹配，内部还含有相关数据(Using different spectral matching algorithms to achieve spectral matching, there are also relevant data inside)

关于奇异值分解，图像特征融合，图像处理的特征图提取方法(About the singular value decomposition )

说明：  小波分析增强滤波。小波变换的原理类似傅里叶变换，只是把三角函数基换成了小波基。与傅里叶变换不同，小波变换有两个变量：scale和translation。scale控制小波函数的收缩，其导数即为频率，translation控制小标函数的平移，平移量对应时间。通过信号的伸缩平移，可以得到某种重合情况，这样积分也会得到一个极大值，不同的是，得到频率成分的同时，还可以知道该频率的时间位置。(Wavelet analysis enhances filtering.The wavelet transform works like the Fourier transform, except that the trigonometric basis is replaced by a wavelet basis. Unlike the Fourier transform, the wavelet transform has two variables: scale and Translation. Scale controls the contraction of the wavelet function, its derivative is frequency, translation controls the translation of the small scale function, and the translation corresponds to the time. By stretching and shifting the signal, you can get some kind of coincidence, so that the integral will also get a maximum, but the difference is that when you get the frequency component, you can also know the time position of that frequency.)