搜索资源列表
MATLBimage-trace
- Even though there are tons of codes from the image registration, artificial intelligence or the robotics community, none of them can easily be used by the strain measuring community. Commercial code is available also and has the advantage of ge
KL
- 基于KL变换的特征提取的方法。 选取数据库中的部分样本（每个人的前5张图片）作为训练样本，其余作为未知的测试样本。从训练样本中得到KL变换矩阵，然后对训练样本和测试样本都进行变换，用变换后的数据作最近邻识别，距离可以为对应灰度值之差的平方和，统计识别率。-KL transform based feature extraction method. Select the database part of the sample (each person' s top 5 pictures)
EdgeDetection
- 边缘检测1.制作一张包括色块、线条的单色背景图片，先对其作低通滤波产生一幅色块和线条边缘模糊的降质图片待处理，采用任意二种边缘检测算法检测待处理图片中的色块和线条的边缘，得到二值化的处理结果图。从结果图中提取色块和线条的边界坐标值，与实际生成原始图像时采用的真实坐标数据作比较，对产生的检测误差作分析。2.拍摄一张包含待检测物体的黑白或者彩色照片，试采用一种边缘检测的算法做待检测物体的自动提取，对生成的结果作分析。-Edge Detection 1. To produce a including
water
- 基于dct变换的数字水印 matlab源码及测试用数据-Based on the DCT Transform Digital Watermarking matlab source code and test data
minipro
- 该系统功能：实现手写识别，能通过对样例库中的数据进行学习，然后能判别、分类新的输入样例。其中包含了Kn近邻算法，贝叶斯参数估计的实现。实现了open test, close test等测试方法。-The system features: realization of handwriting recognition, through the library of sample data for study, and then to identify, classify the new input
Character_Recognition_Training__NN_for_classificat
- 图像特征识别通过神经网络训练方法实现，是学习参考的好资料-you will need first to run the file that name "charGUI4.fig" and on the right side there is a load training set where you have to train the system first, run any data that is should be from 1 to 9 and 0 like ( 1 2 3 4 5 5
H264-DCT-Quanter
- H264中的整数DCT变换编码及量化代码,并附有原文的DCT编码、量化的说明文档。本源码从X264及JM中抽取后整理得来，并用原文档中的示例数据进行了测试，与原文档结果完全一样，是学习H.264不可多得的参考资源. -H264 in the integer DCT transform coding and quantization code, together with the original DCT coding, quantitative descr iption of the doc
gspm
- 自己本机测试最快时60帧/秒,局域网传输只测试一次,结果很让我失望,原因不明~~~~ 说下我的思路过程(我认为过程很重要!) 分块比较+压缩传输与以前那个内存流比较的优缺点: 内存流比较能找到最精确的找到图片差异,但每次都要压缩相同大小的数据,大家可以测试下,那个算法传输瓶颈是压缩算法,最费时间和CPU 分块比较+压缩传输能有效的减少压缩数据量,但缺点也很明显,比较图片的效率肯定没内存流比较高,要加快传输只能找到更好的算法 我选择的是后者,开始的时候是直接把
MIRT2D_WIN
- 自适应正则化方法进行图像配准的matlab代码，带有测试图像数据-Adaptive regularization method for image registration of the matlab code image data with the test
svmTest
- SVM 在JAVA中的运用，简单的DEMO测试，用于做图像识别分析。支援向量机源码调用实例里面有测试资料供简单测试。-The use of SVM in JAVA, simple DEMO test, used to make image recognition analysis. Support vector machine called an instance of source test data for which there is a simple test.
lab_05
- DSP通用Ｉ/O接口控制,系统时钟控制,DAM stop模式视频输出。实现DSP通用接口及按键控制，实现画面分割，内插图像成为ITU-R656 NTSC格式后输出到电视监视器上-DSP Universal I/O interface control, system clock control, DAM stop mode video output. To achieve common interface and button control DSP, based on the complete
Houghcircle
- 基于opencv对采集的图片检测圆,并且将圆上的各点采样,取1024个点,得到这1024个点的灰度值.为后续的FRF变换提供数据.其中灰度的大小代表了此处润滑液的厚度深浅.对工程分析应用非常有帮助.-Opencv image based on the collected test round, and will circle the points on the sample, take 1024 points, 1024 points by the gray values. For the fo
refpaper3_mpcrenglish
- In this paper a stroke feature distribution method is used to develop a scheme for the recognition of alphanumeric characters. The algorithm uses a gray level image of the character as input. The image is normalised. The feature used is strok
paq8n
- PAQ - console archiver with open source developers who jointly have risen in the first place ratings of many test data compression (although at the cost of CPU time and memory).
FaultGenerate
- 输入实验得到的波形，经过单色、二值化处理，得到一系列波形数据，然后重绘波形，并将波形数据输出到控制系统。在测试过程中捕捉故障波形，可以重现测试中的故障，便于调试。-Input waveform experiments, after monochrome, binary, and to obtain a series of waveform data, and then redraw the waveform, and waveform data output to the control sys
fangwentuxiang
- 可以访问图像数据，经测试可以实现，希望大家喜欢-Can access the image data, can be achieved by the test, hope you like
Forecasting-and-residual-test
- 灰色预测和残差检验 可用于预测人口等功能 也可以预测一些数据-Gray prediction and residual test can be used to predict the population of other functions can also predict some of the data
ifft2
- C语言编写的傅立叶正逆变换，经过测试与matlab里面的.m算法结果一致。还有数据测试可用。适用于工程信号处理中。-C language of the Fourier transform, After test and Matlab inside m algorithm results. There are data test available. Suitable for engineering signal processing.
NumberPlateRecognition
- 本程序融合机器学习，利用支持向量机训练车牌样本数据，然后用来识别测试数据，测试结果表明，该训练分类器正确率接近99 -This program fusion machine learning, support vector machine training license plate sample data were then used to identify the test data, test results show that the classifier is trained prop
Test-code-for-SRCNN
- 深度学习，超分辨卷积神经网络，获取测试数据的MATLAB代码-Depth study, the super-resolution convolution neural network, access to test data MATLAB code