Description: 模式识别中的距离分类器C++源代码，并含有一个文字识别的例子，适合模式识别、图像识别开发用！-pattern recognition of distance classifier C source code, and contains an example of character recognition, suitable for pattern recognition, Image Recognition by the development. Platform: |
Size: 1926 |
Author:sjtu |
Hits:
Description: 模式识别中的距离分类器C++源代码，并含有一个文字识别的例子，适合模式识别、图像识别开发用！-pattern recognition of distance classifier C source code, and contains an example of character recognition, suitable for pattern recognition, Image Recognition by the development. Platform: |
Size: 2048 |
Author: |
Hits:
Description: 此代码实现了一个对汉字识别的一个预处理过程，把汉字的笔画按照书写顺序编程数字序列，易变后续处理-This code has a Chinese character recognition as a pretreatment process, the number of strokes of Chinese characters in accordance with the written order of the digital programming sequence, the follow-up to deal with volatile Platform: |
Size: 109568 |
Author:armzm |
Hits:
Description: 思路简要说明：
1、图像二值化
将图片中的各点用0或1表示，1为有效点，0为背景。这里使用的是最大类间方差法
（otsu），在资料中有介绍。
2、去除干扰点
3、分割
将整个的图片分为每个单独的字，在下一步中才能一一识别。
4、与样本库进行对比，寻求最近似匹配
这步是比较核心的地方，由于要识别的图形每次都是随机变化的，我们不能进行完
全匹配识别，所以使用的是‘欧氏距离’来进行最近似匹配，资料中的《自由手写体
数字识别》里面有详细说明。
（样本库文件是按照匹配的特征通过事先编写的程序进行学习得到的）
该识别思路对目前很多验证码有效，识别速度快，正确率基本还可以（在一定程度内
样本量越大正确率越高），不能识别的情况也不少，比如字符粘连，导致程序无法正确
分割，从而识别失败，有朋友介绍神经网络识别方法不错，有空一定要学习下。-A brief description of ideas:
1, image binarization
Pictures of all the points expressed by 0 or 1, 1 for the effective point, 0 for background. Used here is the largest between-class variance
(Otsu), are introduced in the data.
2, removal of interference points
3, partition
The whole picture is divided into each individual character, in the next step in order to identify 11.
4, compared with the sample database, the search might match the recent
This step is to compare the core places, due to the recognition of each of the graphics changes are random, we can not be finished
Identification of the entire match, so the use of the Euclidean distance to be like the recent match, the information in the "free handwritten
Digital Identification, "which is described in detail.
(Sample library features in accordance with the matching pre-prepared through the process of learning to be)
The identification of a lot of ide Platform: |
Size: 874496 |
Author:yangq |
Hits:
Description: 本文提出了一种新的车辆许可证盘子识别,并在此基础上提出了一种自适应图像分割方法-In this paper, a new algorithm for vehicle license
plate identification is proposed, on the basis of a novel adaptive
image segmentation technique (Sliding Windows) in
conjunction with a character recognition Neural Network. The
algorithm was tested with 2820 natural scene gray level vehicle
images of different backgrounds and ambient illumination.
The camera focused on the plate, while the angle of view and
the distance from the vehicle varied according to the
experimental setup. The license plates properly segmented
were 2719 over 2820 input images (96.4 ). The Optical
Character Recognition (OCR) system is a two layer
Probabilistic Neural Network with topology 108-180-36, whose
performance reached 97.4 . The PNN was trained to identify
multi-font alphanumeric characters from car license plates
based on data obtained from algorithmic image processing. Platform: |
Size: 831488 |
Author:keithe |
Hits:
Description: 自己制作基于“欧氏距离的算法”来识别文字的相似性，从而来识别手写文字的程序，开发环境是matlab.需要讲手写的字加到字库才可以哦。-Produce their own based on the " Euclidean distance algorithm" to identify the similarity of the text, handwritten text in order to identify the procedures, the development environment is matlab. Need to speak the word handwritten character can only be added to Oh. Platform: |
Size: 35840 |
Author:mr hai |
Hits:
Description: A Handwritten Character Recognition System Using Directional Element Feature and Asymmetric Mahalannobis Distance,A Handwritten Character Recognition System Using Directional Element Feature and Asymmetric Mahalannobis Distance Platform: |
Size: 143360 |
Author:zcq |
Hits:
Description: 文字和字符识别，各种识别的样本生成，与生成样本的图片同距同分辨率拍摄下来识别。运动方式下角度考虑与样本所用相同的分辨率和拍摄距离的图片识别，对运动物体图像通过不断检测所标志的区域特征来后定点提取图像识别，对固定不同距离的方式通过缩放处理方式后再识别.针对不同应用也方便在代码里进一步完善-Text and character recognition, various identification of the sample to generate the same to generate sample image from the same resolution filmed identification. Viewpoint movement pattern and the sample with the same resolution and the shooting distance picture identification characteristics of the marked areas of the moving object image by continuously detecting sentinel extracted image recognition, the fixed different distances through the scaling processing mode and then identification is also convenient for different applications to further improve in the code Platform: |
Size: 678912 |
Author:曾祥园 |
Hits:
Description: 以数字与字母识别系统的基本处理流程为主线，从待识别数据的获取入手，通过预处理、特征的提取与选择，到分类器的设计等部分都进行了较为详尽的分析与研究，MATLAB仿真实验表明；采用最小距离法对所给出的一组数字及字符图片进行不同的分块识别，最终得出分8块识别率为85.71 ；分16块识别率为95.71 ；分20块识别率为95 ；具有较高的识别率。-The basic process flow of the numbers and letters recognition system as the main line, starting from the acquisition of the data to be identified, preprocessing, feature extraction and selection, to the part of the classifier design a more detailed analysis and research, MATLAB simulationexperiments show minimum distance given a set of character image. different sub-block identification, and ultimately arrive at 8 recognition rate of 85.71 16 points recognition rate of 95.71 20 identifyrate of 95 has a higher recognition rate. Platform: |
Size: 4096 |
Author:刘下 |
Hits:
Description: 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。
对字符正确识别之后，用事先对对每一个字符的录音根据对应字符顺序播放。在对车牌区域识别出错、字体分割出错时程序暂停，并有语音提醒。-The source is based on MATLAB license plate recognition and voice broadcast. The system for the small family car blue and white license plate recognition. Approximate positioning of the blue color candidate area based on the ratio of RGB images. However, due to the RGB color space, the Euclidean distance between two points is not a linear scale and color distance, in the blue region of the positioning range setting is not well controlled. Therefore, the positioning error is the main cause. This appears in the picture under a blue background more recognition rate will drop, can not effectively extract the license plate area. This paper presents an adaptive adjustment program. On the segmented regions to identify adjustments. According to the aspect ratio, blue and white area than the candidate repeatedly positioning. Finally found the license plate area. After correctly identify the characters, prior to use for recording each character according to the corresponding sequence of characte Platform: |
Size: 24704000 |
Author:longfei |
Hits:
Search - distance recognition character