资源列表
regnize
- 实现0-9个手写数字的识别 使用了多种方法:最近邻、势函数、神经网络、贝叶斯分类器-To achieve recognition handwritten digits 0-9 using a variety of methods: the nearest neighbor, potential function, neural networks, Bayesian classifier
Adaboost
- 总结和整理了一些adaboost文献资料 对做人脸检测和识别有一定的帮助-Adaboost summarized and compiled some literature on the human face detection and identification of some help
opencv-biaopan
- opencv 1.0下表盘识别的程序 识别出表盘的指针刻度-opencv 1.0 identification procedures under the dial pointer to identify the scale of the dial
BP
- 本算法利用BP算法处理手写体识别,为宜较为经典的机器学习算法-The algorithm uses the BP algorithm processing handwriting recognition, appropriate more classic machine learning algorithms
ExtractGLCM
- 灰度共生矩阵方法用来提供一幅图像的灰度共生矩阵值,然后送到分类器中进行分类-gray level co-occurrence matrix,a very important method of extracting features
face-detection-designed-by-matlab
- 利用matlab实现人脸识别,将facefind函数生成.dll动态链接库,检测准确,可以检测到侧面图像,识别率95 以上。-Using matlab to achieve recognition, the facefind function generates. Dll the dynamic link library, testing is accurate, can detect the side of the image, recognition rate of 95 or more.
Source
- c#实现的人脸识别程序,请参考!适合搞人脸识别的朋友-A FER program is implemented by C#,Please use it,It is suitable to you if you learn FER field!
License-plate-localization
- 很好的车牌识别源码,用VC实现的,值得初学者学习和借鉴,-Good source license plate recognition, with the VC to achieve, beginners should learn from,
ISODATA
- ISODATA算法是一种基于统计模式识别的,非常经典的非监督学习动态聚类算法,有较强的实用性。ISODATA算法不仅可以通过调整样本所属类别完成样本的聚类分析,而且可以自动地进行类别的“合并”和“分裂”,从而得到类数比较合理的聚类结果。-ISODATA algorithm is based on statistical pattern recognition, unsupervised learning is the classic dynamic clustering algorithm
Handwritten-recognition
- 手写数字识别程序。可用于对数字,英文,汉字的识别-Handwritten numeral recognition program. Can be used for digital, English, Chinese character recognition
Feature-Extraction
- 特征提取与图像处理(第2版)<尼克松(Mark S.Nixon) 英文版 适合研究图像匹配、目标识别等方面的人员使用。-Feature Extraction and Image Processing (2nd Edition) (Mark S. Nixon)。 Specially for who study on image pattern,tracking。
OCR-AD--2.01
- OCRAD 的文字识别系统,英文识别效果相当不错!只支持引文识别,这个比tesseract差一点。-OCRAD text recognition system, recognition results have been very good English! Citation only support recognition, the ratio was almost tesseract.