搜索资源列表
ofdm_1024
- OFDM系统仿真，采用1024子载波，内含卷积编码，交织，QPSK调制，上、下采样滤波，信道估计等
DCTJPEG
- 使用Matlab编程实现发送端信源编码（图像压缩）和接收端信源解码（图像重建），其中的编解码算法可采用基于DCT变换的JPEG图像压缩标准；测试图像加噪，然后对去噪后的图像进行压缩与重建，分析压缩性能；研究基于下采样和内插的图像压缩算法；可将8bit/pixel灰度图像扩展至24bit/pixel彩色图像，或者采用无损压缩算法对图像压缩。-Sender using the Matlab programming source coding (image compression) and the r
nsct_toolbox
- NSCT非下采样Contourlet变换，实现对图像的多尺度多方向分解，是一种真正的二维图像表示方法-NSCT nonsubsampled Contourlet transform, to achieve multi-scale multi-orientation image decomposition, is a true two-dimensional image representation
nsct
- 基于非下采样轮廓波的边缘提取程序。最新的程序-Under the sample based on non-profile wave of edge detection procedure. The latest program
生成低分别率图像程序
- 利用成像模型，由高分辨率图像生成位移、下采样图像。用于高分辨率图像重建中低分辨率图像仿真
ddc_30m
- 中频70M，30M带宽LFM信号，采样率为102.4M，数字下变频后，还进行了三倍抽取，最后还得到I，Q两路信号 -IF 70M, 30M bandwidth LFM signal, the sampling rate of 102.4M, digital down conversion, the samples were carried out three times, the last also I, Q two-way signal
BPSK
- 上采样和下采样的matlab程序，以函数形式给出，有三种模式。-upsampling,downsampling,matlab,three mode
OFDM-qpsk
- OFDM，QPSK调制下仿真程序，包括信道编解码，调制解调，上下采样-OFDM, QPSK modulation simulation program, including channel coding and decoding, modulation and demodulation, the upper and lower sampling
filter-deniose
- matlab下实现降噪，功率谱计算。％程序中设定采样点数为256个，采样频率为10000hz， ％输入信号为随机噪声和两个正弦的合成信号 ％设定的滤波器的截止频率为3500hz ％通过挈比雪肤滤波器，运行程序，比较滤波前后的频域波形 ％滤波去除了高频信号-matlab achieve noise reduction, power spectrum calculation. Sampling procedure set for 256 points, the sampling f
Contourlet
- 非下采样轮廓波变换的应用，可以对该变换有一个很好的理解-Transform under the profile of non-sampling applications, the transformation can have a good understanding of
nsct_toolbox
- 对图像进行非下采样小波变换，得到一个低频子带和若干个高频子带。-Under the image of non-sampling wavelet transform, to get a low-frequency sub-band and a number of high-frequency sub-band.
mycontourlet
- 基非下采样的Contourlet变换的图像融合，要放到NSCT工具包中执行-Under the base of non-sampling Contourlet transform image fusion, to put the implementation of the toolkit NSCT
Copy-Detection
- 使用下采样帧构成的列向量的PCA系数作为每帧特征（原则上，也可以使用其他类型的关键帧特征），多帧特征构成特征矢量，以高斯概率模型的后验概率衡量相似性、并建立k维树结构的索引进行搜索，文中对噪声、模糊、再压缩、加Logo，及其这些变换的两两组合进行了实验（丢帧很少时也可以），速度很快，适合于大规模因特网视频的搜索。-Use sampling frame column vector of the PCA coefficients as each frame features (in principl
wave
- 基于非下采样小波变换的纹理图像特征提取程序-The sampling based on the wavelet transform of texture image feature extraction
Daubechie_wavelet_ImageRecover
- 采用Daubechie小波，构建正交矩阵，对bmp图像进行小波变换后的基本重建恢复，需要下采样，压缩为64*64以下的图像，重构图像与原图基本一致。-Daubechie wavelet to construct the orthogonal matrix, bmp image to restore the basic reconstruction after wavelet transform, sampling, compressed into less than 64* 64 images,
xiacaiyang
- 将高清1920x1080YUV视频序列按2：1比例下采样-The high-definition video sequences 1920x1080YUV downsampling ratio of 2:1
ofdm_2
- OFDM仿真包括信号源,卷积码编码，交织，qpsk调制，IFFT，组帧，上采样，通过FIR滤波器 通过高斯信道，再经过低通FIR滤波器，下采样，拆帧，FFT，qpsk解调，接收数据；计算误比特率-OFDM simulation including the signal source, sampling convolutional encoding, interleaving, QPSK modulation, IFFT, frame,, through the FIR filter throu
FeaturExtractionCode
- 提取自然图像的区域特征，先是对图像进行下采样进行降维，然后彩色图像三通道对每一个通道进行提取颜色信息。-The regional feature of the natural image is extracted, and the image is down-sampled, and then the color image is extracted for each channel.
CNN
- 一个卷积层+一个下采样+softmax实现mnist识别(implement a simple CNN)
CNN Matlab代码
- 利用大量图像数据对卷积神经网络算法进行训练，通过卷积、池化、下采样以及全连接层训练后的卷积神经网络在图像识别精度越来越高。(By using a large number of image data to train the convolutional neural network algorithm, the accuracy of the image recognition is higher and higher by convolution, pooling, down sampling