资源列表
iris
- iris数据分类,bp神经网络实现,适当调整训练次数和训练精度,分类准确率可达99 以上-iris data classification, bp neural network, training times and training appropriate adjustments to the accuracy, classification accuracy rate of 99 or more
rgb
- 读取并分解24位真彩图像,可以提取其中数字的特征值并输出,也可输出分解后的图像-Read and decomposition of 24-bit true color image, you can extract the characteristics of digital values and output, can output the image decomposition
SLLE
- 稀疏流形建模及其在人脸识别中的应用论文介绍稀疏理论 以及sparse LLE -Sparse manifold modeling and its application to face recognition paper introduces the theory and sparse LLE sparse
violaJones_CVPR2001
- 该论文为Viola Jones 的经典人脸识别系统的设计思路,以及其在应用中的检测率和假阳性率-The paper is a classic Viola Jones face recognition system design and its application in the detection rate and false positive rate
Relief
- 经典特征选择程序,在特征提取完成后进行特征选择可以达到提取有用成分的目的-Feature selection procedure, after completion of the feature extraction feature selection can achieve the purpose of extracting useful components
FaceRecog_src
- 程序结构 整个工程可以分为3个部分:算法、功能和应用。 算法部分 算法部分目前分为4个模块:人脸对齐、光照归一化、特征提取和选择、子空间降维,每个模块是一个项目,每个项目生成一个dll供功能部分显式调用。 功能部分 功能部分只有一个项目FaceMngr,该部分依赖于算法部分,实现人脸注册、训练、识别、导入/导出等具体功能。该项目生成一个dll供应用部分隐式调用。 应用部分 人脸识别Demo. 另外,工程中
pilaojiashi
- 在VC++环境下利用OPENCV实现人眼检测和识别,并判断驾驶员是否疲劳,自动提醒驾驶员停止驾驶。-In VC++ environment using OPENCV achieve detection and recognition of the human eye and determine whether driver fatigue, and automatically alerts the driver to stop driving.
DCT
- 先用小波变换进行降维后,再用DCT进行特征提取,然后用SVM分类识别,SVM需先安用libsvm工具箱,然后再可以运行,该程序包含ROL人脸库,一并上传。-First reduce the dimension of the wavelet transform, the then DCT feature extraction, and then use SVM classification, SVM must be safe to use libsvm toolbox, and then you
hr
- 人体动作识别,用于智能视频分析开发资料,主要应用于DSP嵌入式-Human motion recognition for intelligent video analysis, development of material, mainly used in DSP embedded
PCAKPCA
- PCA和KPCA程序,matlab实现,可用于模式识别时做降维或特征提取处理-PCA and KPCA program, matlab implementation, pattern recognition can be used to do when dealing with dimensionality reduction or feature extraction
motempl
- 基于OPENCV的图形识别与跟踪,运动检测程序,用于提取运动物体轮廓,并跟踪其运动路线-OPENCV motion detection procedures and track their movement routes
codereg
- 简单验证码识别,几年前写的,拿出来分享,首先2值话,然后像素对比识别-Simple verification code recognition, written a few years ago, to share out the first two values, then, and then identify the pixel contrast