搜索资源列表
VCPPCODErecognize
- 基于神经网络的文字识别系统 本目录主要包括:文档和两个源代 其中源代码是识别程序的,另一个是矩阵类库的。已经保存了训练好的网络权值,所以第一次使用识别程序时,可以直接运行Release目录下的可执行文件,然后对图片目录中的测试图片进行读入、识别。当然也可以自行用训练样本训练网络,不过要注意训练样本的选择,否则可能识别率很低。训练样本选择的原则是,尽可能的有代表性,在训练时间不至于太长的情况下训练样本数目尽量多。 -Neural network character recognitio
voice-recognition
- 实现基于MFCC和HMM及Viterbi的语音识别系统,识别率可达到80 以上-MFCC and HMM-based speech recognition systems and Viterbi recognition rate can reach 80
carc
- 车牌识别的完整程序,包含模板和测试图片,识别率较高-Complete license plate recognition program, including templates and test images, the recognition rate is higher
plate-identification
- 针对夜间光线不足导致对车牌识别率下降,该代码能对夜间机动车车牌图像进行增强与识别。-Pattern Recognition
ocrrec
- 这次还是课后作业 做的是hopfield的手写体字符识别 识别率对库文件的1能达到90 -This is homework to do is hopfield handwritten character recognition recognition rate for a library file can reach 90
carocr-
- 这是一款最新的车牌识别系统,可识别各种车型,刮风下雨,出太阳,识别率仍很高!-This is a new license plate recognition system that can identify a variety of models, wind and rain, the sun was still high recognition!
Plate-recognition-system-source-code
- 本源码是基于MATLAB实现车牌识别并语音播报。 本系统针对家庭小型车蓝底白字车牌进行识别。根据彩色图像的RGB比例定位出近似蓝色的候选区域。但是由于RGB三原色空间中两点间的欧氏距离与颜色距离不成线性比例,在设定蓝色区域的定位范围时不能很好的控制。因此造成的定位出错是最主要的。这样在图片中出现较多的蓝色背景情况下识别率会下降,不能有效提取车牌区域。对此本文提出了自适应调节方案。对分割出来的区域进行识别调整。根据长宽比,蓝白色比对候选区域进行多次定位。最终找到车牌区域。 对字符正确识别之
TTS
- 自己写的语音识别系统,可以调语速。识别率很高,希望对大家有所帮助-Write your own voice recognition system that can adjust the speech rate. High recognition rate, we hope to help
SinVoice_plus
- 目前声波通信已经在iphone和android中广泛的应用起来了,涉及到数据和文件传输,以及支付等众多领域。比如iphone中的chirp,android中的茄子快传,支付宝的声波支付,小米快传等。该声波通信程序是商用改进版本,在前一个开源版本(SinVoice)的基础上,做了许多优化: 优化如下: * 1. 识别效率更高,几乎达到100 ,完全可以达到商业用途标准,比chirp,支付宝,茄子快传等软件的识别效率更高。 * 2. 能支持更多复杂场景的识别,在有嘈杂大声的背景音乐,
face-recognition-based-on-pca_matlab
- 这是基于主成分分析的人脸识别算法,包含下采样,特征脸,数据字典等;识别率很高,希望对做识别这块的初学者有帮助-This is a face recognition algorithm based on principal component analysis, including down-sampling, face features, data dictionaries very high recognition rate, I hope to do this is to help beg
PSKmodulation
- 提出了一种不仅计算量小,而且可以在低信噪比下正确识别相位调制信号的方法。利用Matlab工具进行仿真,仿真结果表明,在信噪比不低于2 dB的情况下,对PSK子类调制信号的识别率可以达到93 以上。 -Proposed a method of calculating only a small amount, and can correctly identify the phase modulation signal at a low SNR. Use Matlab tool for simul
Face-recognition-
- 人脸的研究是跨越人文科学与自然科学的新兴交叉研究领域,在最近几年得到了模式识别领域众多学者的重视,也取得了良好的研究成果。针对人脸这种生物特征的识别,在其过程中最为重要的一个环节是特征的提取,更好的提取出人脸的特征,将会使得识别更有有效和准确,提高分类的同时,也是的识别率有良好的提高。-Face recognition research is a new cross field across the humanities and natural science, had many scholar
vq-Speakerrecognized
- vq矢量量化,对说话人进行识别,内置有语音样本,识别率高达90 。对语音信号进行读取,预处理,特征参数提取,向量量,训练,对比识别。-vq vector quantization for speaker identification, built-in voice samples, the recognition rate of 90 . Read the speech signal, preprocessing, feature extraction, the amount of vector
ZXingDemo
- 基于Android的二维码扫描Demo,识别非常快,识别率很高-Android-based two-dimensional code scanning Demo, recognition is very fast, very high recognition rate
hf
- 使用加强版霍夫变化,使得车道线识别率百分之百,不存在车道线断点。属于原创。-Using an enhanced version of the Hough changes, making lane hundred percent recognition rate, there is no lane breakpoint. Belong to the original.
aa
- 基于线性矩阵分析的盲识别技术源代码,可以对任意线性分组码进行全忙识别,并在一定信噪比下统计识别率-Blind analysis based on linear matrix recognition source code, can be arbitrary linear block code to identify all busy, and some SNR statistical recognition rate
Handwritten-Character
- 基于CNNs的手写字符识别系统,载入MNIST手写字符数据库,通过训练提取特征,达到99 的识别率-Based on CNNs handwritten character recognition system, load MNIST handwritten character database, extract features through training, up to 99 recognition rate
hc
- 手写计算器,可以实现手写加减乘除,识别率挺高。matlab2010b及以上版本打开-Handwriting calculator, you can achieve a handwritten addition, subtraction, recognition rate pricey. matlab2010b and later opened. .
ImageRecognition
- 智能识别,很基础的东西,可以识别不是很复杂的验证码,识别率还好-Intelligent recognition, very basic things, you can identify verification code is not very complex, but fortunately, the recognition rate
SVM_pedestrian-identification
- 利用SVM机器学习对视频中行人进行识别。本工程的识别率较高。如有不足之处,请大家多多指点。-Using SVM machine learning of pedestrian to identify in the video. The project has higher recognition rate. If there are any shortcomings, please advice for me.