Bidimensional Empirical Mode Decomposition Applications in Signal and Image Processing

Bidimensional empirical mode decomposition (BEMD) is an adaptive, data-driven framework for analysing two-dimensional signals and images by decomposing them into a set of bidimensional intrinsic mode functions (BIMFs). Extending the one-dimensional EMD algorithm to higher dimensions, BEMD iteratively sifts local extrema surfaces to extract geometrically meaningful components across scales. This multi-scale decomposition facilitates the isolation of textures, edges, noise and structural patterns without requiring prior basis functions. In signal processing, BEMD provides enhanced denoising and feature-extraction capabilities for non-stationary, non-linear data streams, while in image processing it has been applied to speckle noise suppression, texture segmentation, face recognition under varying illumination, medical image segmentation and surface topography characterisation. The ability of BEMD to preserve fine details and adapt to the intrinsic content of each signal renders it suitable for diverse real-world applications, from synthetic aperture radar noise reduction to real-time biomedical signal analysis. Recent advances have focused on overcoming computational challenges such as the edge effect, accelerating decomposition through parallel or hardware-accelerated architectures, and integrating BEMD with machine-learning techniques such as kernel principal component analysis and analytic phase theory. These developments highlight BEMD’s broad potential in both fundamental research and practical deployment across scientific and engineering domains.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Bidimensional Empirical Mode Decomposition Applications in Signal and Image Processing publication trend

The graph below shows the total number of articles in bidimensional empirical mode decomposition applications in signal and image processing across all publications each year (not limited to Nature Index journals).

Technical terms

Bidimensional Empirical Mode Decomposition (BEMD): A method that decomposes two-dimensional signals or images into adaptive components (BIMFs) by sifting local extrema surfaces without predefined basis functions.

Bidimensional Intrinsic Mode Function (BIMF): A single, two-dimensional component produced by BEMD that represents oscillatory modes in an image or signal at a given scale.

Edge effect: Distortions and artefacts occurring at the boundaries of data during decomposition, often requiring special extension or mirror techniques to mitigate.

Speckle noise: Granular interference commonly found in coherent imaging systems (e.g. SAR), which degrades image quality and complicates feature extraction.

Kernel Principal Component Analysis (KPCA): A non-linear dimensionality-reduction technique that maps data into a high-dimensional feature space for principal component analysis, often combined with BEMD to enhance decomposition robustness.

References

1. Face Recognition under Varying Illumination Using Green’s Functionbased Bidimensional Empirical Mode Decomposition and Gradientfaces. ITM Web of Conferences (2016).
2. Distance regularized level set evolution in magnetic resonance image segmention based on bi-dimensional ensemble empirical mode decomposition. Acta Physica Sinica (2016).
3. SAR image noise suppression of BEMD by the kernel principle component analysis. IET Image Processing (2020).
4. A novel parallel unsupervised texture segmentation approach. Discover Applied Sciences (2023).
5. Relative vibration identification of cutter and workpiece based on improved bidimensional empirical mode decomposition. Frontiers of Mechanical Engineering (2020).