object_tracking_based_on_kalman

基于BoVW的图像特征提取，以及图像检索(Image feature extraction and image retrieval based on BoVW)

说明：  这是一个比较实用的ADBOOST算法的源代码，它是比较准确，希望大家喜欢。(This is a more practical algorithm ADBOOST source code, it is more accurate, I hope you like.)

完成了对车道线进行了识别，主要流程为灰度化，边缘检测，hough变换。(We completed the lane line identification, the main process of graying, edge detection, hough transform.)

四种滤波（SVD滤波，PCA滤波，DCT滤波，DWT滤波）对图像处理，有程序代码和对应图片(Four kinds of filter for image processing)

用于虹膜定位……定位外圆之前需要进行边界跟踪，此程序实现此功能(For Iris Location ... Location ... before the need for outer border tracking, this procedure realize this feature)

分割点云，并通过局部拟合求取法向量，实现点云压缩(Split point cloud, and through local fitting strike normal vector of point cloud compression)

说明：  本算法是matlab代码，通过icp点云匹配算法实现三维点云的配准(The algorithm is matlab code, and the registration of three-dimensional point cloud is realized by ICP point cloud matching algorithm.)

改进的分水岭算法，能够较好地提取对象轮廓、准确得到物体边界的优点。在MATLAB6.5下顺利编译通过(improve the watershed algorithm can be used to extract object contours, to be precise object boundary advantages. In MATLAB6.5 smoothly through the compiler)

image decomposition and reconstruction

实现RGB颜色空间到HSV颜色空间变换的函数。(achieve RGB color space to the HSV color space conversion function.)