BEMD-Based Image Decomposition Using Delaunay Triangulation and Cubic Interpolation
- Login to Download
- 1 Credits
Resource Overview
BEMD decomposes images through envelope construction via Delaunay triangulation and cubic interpolation, producing IMF components and residual outputs. Input: grayscale BMP image. Output: IMF1, IMF2, IMF3, and residual components for advanced image analysis.
Detailed Documentation
This algorithm employs Bidimensional Empirical Mode Decomposition (BEMD) to extract multidimensional information from images through decomposition. The envelope construction process utilizes Delaunay triangulation for surface interpolation, followed by cubic spline interpolation to generate smooth upper and lower envelopes. Key implementation steps include: 1) Identifying extrema points across the image matrix, 2) Constructing triangular meshes using Delaunay's algorithm, 3) Applying cubic interpolation to form multidimensional envelopes, and 4) Iterative sifting to obtain Intrinsic Mode Functions (IMFs). The final output consists of three IMF layers (IMF1, IMF2, IMF3) representing high-to-medium frequency components, and a residual layer containing low-frequency image features. These components enable advanced applications in image enhancement, texture analysis, and multiscale feature extraction.
- Login to Download
- 1 Credits