搜索资源列表
文字识别程序
- 目前该手写体识别系统主要分为 预处理模块： 主要包括训练数据和识别数据的读取，归一化，二值化 特征提取模块：主要包括笔划方向特征和网格密度特征，还可以根据对识别率的要求继续增加其他特征 识别（分类器）模块：主要包括SVM方法和BP神经网络的方法，其中SVM方法的识别率较高，SVM＋网格密度特征， 在小字符集情况下，达到了识别率97％以上 采用OO思想编写，适合做二次开发-currently the handwriting recognition system consists of pretre
手写识别系统
- 数字识别，识别手写体。 使用神经网络算法编写。 识别率一般，带训练功能-Number recognizing system,used to recogize the hand writing figures.It is writen by nero network and has low recognizing effects but we can exercise it.
szsbxt
- 基于神经网络的数字识别系统，先要训练才能识别，每识别一种字体的数字要进行相应的训练，具体看说明。-based on neural network digital identification system must be trained to identify, each font identification of a number corresponding to the training, specifically looking for.
recoginition_number_classic
- 对有噪声的数字 利用神经网络进行识别， 训练效果显著-right is the number of noise using neural network identification, training results significantly
BPl
- 一个基于BP神经网络的matlab程序，可以实现对几种字体0-9的数字识别，这个文件训练网络的压缩包-a neural network based on BP's Matlab procedures, Some can be achieved on the font 0 -9 identification number, the document training network compressed
angela-work
- 运用bp神经网络方法实现对8*16像素的手写体0~9数字识别的代码，列出了训练样本以及测试样本-use bp neural network method of 8 * 16 pixels handwritten 0 ~ 9 digital identification code, listed training samples and test samples
nn
- 神经网络训练，可参考相应的文章，杜达《模式分类》等
20070905
- 数字识别之神经网络法 数字训练、识别，功能强大
Matlab实现基于神经网络的文字/字母识别
- 程序运用于MATLAB，实现了神经网络的文字识别，目前是对英文字母，当然对汉字要难些，而且可能会不是特别准确（因为需要训练的样本多了）。
RbfSingleCharRecognition
- 径向基神经网络的字符和数字识别程序，在训练的基础上进行识别-RBF neural network and digital character recognition program, the training conducted on the basis of identification
Classification
- 模式分类。包括:训练样本设计、模板匹配分类器、Bayes分类器、线性函数分类法、非线性分类法、神经网络分类法-Pattern classification. Include: training sample design, template matching classifier, Bayes classifier, a linear function of classification, non-linear classification, neural network classificat
DigitalRec
- 手写体数字识别的VC实现 使用神经网络算法对手写体数字进行识别，训练后识别率可达90％左右。-Handwritten Digit Recognition of the VC to achieve the use of neural network algorithm for handwritten numeral recognition, training, recognition rate can reach about 90.
shibie
- 基于bp神经网络的车牌识别系统 神经网络部分训练有待提高-The design of intelligent transportation systems for applications such as electronic toll,traffic flow analysis,speed Limit and red light violation enforcement has been attracted an increasing attention from researchers.I
BPneuralnetworkbasedonthematlab
- 一个基于BP神经网络的matlab程序可以实现对几种字体0-9的数字识别这个文件训练网络的压缩包-BP neural network based on the matlab program can achieve several fonts 0-9 to identify the document number of the compressed packet network training
Recog
- 用于数字字符识别。包括对含有数字的图片进行二值化，去噪，分隔字符，归一化处理，训练神经网络并识别。-For digital character recognition. Including the number of pictures that contain binary, denoising, separation characters, normalization, and identification of training neural networks.
ann数字识别
- 首先训练一组文字，然后进行识别，识别准确率较高，比较容易上手的代码(First, train a set of words and then identify them)
character recognition
- CNN卷积神经网络对字符进行训练、字符识别相关代码(Printed ink jet character recognition)
□…22 LVQ篦□±□璱□蓄蓄□鉲礥楗辛謇
- BP神经网络matlab完整代码，包括训练代码、测试代码(BP neural network matlab complete code, including the training code, test code, the image directly into the network is trained to)
cnn
- 基于python tensorflow框架构建的卷积神经网络用来识别图像，附带训练数据集的制作代码。(The convolution neural network based on the python tensorflow framework is used to identify images with the production code of the training data set.)
shoushishibie
- 进行手势识别 内涵简单的训练图片和代码 含有图像分割函数(Gesture recognition Concise training pictures and codes Contains image segmentation functions.)