Wavelet Toolbox 为信号的时频分析和图像的多尺度分析提供 App 和函数。您可以对数据进行去噪和压缩，并检测异常、变化点和瞬变。该工具箱通过提供时频变换和自动特征提取，包括散射变换、连续小波变换（尺度图）、Wigner-Ville 分布和经验模态分解，来支持以数据为中心的人工智能 (AI) 工作流。您可以使用小波、小波包和剪切波变换从图像中提取边缘和方向特征。
这些 App 支持您以交互方式执行时频分析、信号去噪或图像分析，并生成 MATLAB 脚本来重现或自动化您的工作。
您可以从工具箱函数中生成 C/C++ 和 CUDA® 代码以用于嵌入式部署。
加速和部署
使用 GPU 和多核处理器运行支持的函数，从而加快代码执行速度。对于支持 C/C++ 代码生成的 Wavelet Toolbox 函数，可使用 MATLAB Coder 为其生成独立的 ANSI 兼容 C/C++ 代码。为支持的函数生成优化 CUDA 代码以在 NVIDIA® GPU 上运行。
产品资源：
Wavelet Toolbox FAQs
Wavelet Toolbox provides apps and functions for time-frequency analysis of signals and multiscale analysis of images, enabling you to denoise and compress data, detect anomalies, change-points, and transients.
You can perform time-frequency analysis using continuous wavelet transforms (scalograms), Wigner-Ville distribution, empirical mode decomposition, and extract edges and oriented features from images using wavelet, wavelet packet, and shearlet transforms.
The toolbox enables data-centric AI workflows by providing time-frequency transforms and automated feature extraction, including scattering transforms, which can generate 2D time-frequency maps for use as inputs to deep convolutional neural networks.
Yes, the Wavelet Signal Denoiser app lets you visualize and denoise 1D signals, and wavelet techniques can denoise images while retaining features that other denoising techniques might remove or smooth.
Yes, you can generate C/C++ and CUDA code from toolbox functions for embedded deployment using MATLAB Coder and speed up your code using GPU and multicore processors for supported functions.
The CWT allows you to analyze signals jointly in time and frequency with visualization available through scalograms in the Time-Frequency Analyzer app.
Discrete multiresolution analysis uses the decimated discrete wavelet transform (DWT) to analyze signals, images, and 3D volumes in progressively finer octave bands, and can decompose nonlinear or nonstationary processes using empirical mode decomposition.
Yes, you can use wavelet and wavelet packet algorithms to compress signals and images by removing data without affecting perceptual quality.
