资源列表
XIHUA
- 本程使用c++开发语言编写的，功能是将bmp图像进行细化处理，比如，可以将比较粗的汉字细化成较细的汉字。-This routine using c++ development language, the function is to BMP image thinning processing, for example, can be more crude refining the Chinese characters into a fine characters.
Nonlinear_Diffusion_v1.2
- 代码支持齐次线性和非线性（总变差和边缘增强流动）的任意尺寸领域各向同性扩散（标量/灰度图像，彩色图像和矩阵向量/结构张量）。添加剂算子分裂（AOS）以及高斯正则化的实现加速计算。-The code supports homogeneous and linear and nonlinear (Total Variation and Edge Enhancing flow) isotropic diffusion of arbitrary dimensioned fields(scalar~gray
QingXie
- 使用c++编写的程序，功能是将文本图像中的文字切分出来。-It is c++ program, the function is to segmente the text in the image .
HangZiQieFen
- 使用c++编写，功能是将文本图像中的字切分出来。-The program used c++ language,and the function is segment the word in the text image .
disk
- 一个散焦模糊自动复原的算法，自动估计散焦半径，在频率域识别参数，并用最小二乘滤波复原-A defocus blur automatic recovery algorithm automatically estimated defocus radius, in the frequency domain identification parameters, and by the least squares filter restoration
mrw
- Daubachies basis function for matlab wavelet multiresolution
ViBe_track
- 利用opencv，基于vibe算法，编写的前景提取程序，效率高，速度快，效果好-Using OpenCV, based on the vibe algorithm, the preparation of the prospect of the extraction process, high efficiency, high speed, good results
ObjectPoseEstimation-master
- 物体的姿势识别 效果不错 可以下载试试看-Object recognition of the effect of good posture can be downloaded to try
PartsBasedDetector-master
- 人体姿势识别 可以识别出人体的14个节点 连接节点然后得出姿势图- 人体姿势识别 可以识别出人体的14个节点 连接节点然后得出姿势图 Human body posture recognition can identify the body of the 14 nodes connected to the node and then draw a picture
pf-segment.tar
- 人体分割 效果很好 将图片分割得到人体特征-Human body segmentation effect is very good image segmentation
gco-v3.0.tar
- 图像分割优化 能有效优化分割图片 可以下载-Image segmentation optimization can effectively optimize the image can be downloaded to try to download
shibie
- 手写体数字的识别，采用bp神经网络，有很好的效果-Recognition of handwritten digits, using bp neural network, the results were OK