Hilbert-Huang Transform System for Real-Time 2-D Image Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Hilbert-Huang Transform (HHT) data processing systems are limited to one-dimensional data and are not real-time, making them inadequate for processing large two-dimensional images, which requires significant computational resources and is computationally complex, especially for images of size N=2048.
Innovation Solution
A real-time Hilbert-Huang Transform data processing system with 2-D capabilities is developed, utilizing a computer program to construct and smooth upper and lower envelopes for input images based on pixel window parameters, and employing a method to reduce computational complexity from O(N^4) to O(N^3) by establishing separability and using parallel processing for Empirical Mode Decomposition, allowing for efficient processing of large images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If HHT is applied to 2-D images, then spectral analysis capability for non-linear and non-stationary sources is improved, but computational complexity increases from O(N^3) to O(N^4)
Solution Approach 1:
The patent divides the 2-D image processing into multiple 1-D processing stages. First, the image is decomposed into horizontal and vertical components, then HHT is applied separately to each dimension. This segmentation reduces the computational complexity from O(N^4) for direct 2-D HHT to O(N^3) by processing dimensions independently and combining results.
2Speed
If standard HHT-DPS is used for 1-D data, then processing speed is maintained, but 2-D image processing capability is lost
Solution Approach 1:
The patent creates a universal processing framework that can handle both 1-D signals and 2-D images using the same core HHT algorithm. By implementing separability, the system maintains the efficient 1-D processing speed while gaining 2-D image processing capability, making the system multi-functional without requiring separate algorithms for different data types.
3Loss of time
If real-time processing is implemented, then processing time is reduced, but computational accuracy may be compromised
Solution Approach 1:
The patent performs preliminary data preparation and preprocessing steps before applying the full HHT algorithm. By organizing data structures in advance, pre-computing certain components, and preparing the signal decomposition framework beforehand, the system reduces real-time processing time while maintaining computational accuracy through careful preservation of all necessary signal characteristics.
Data Source
AI summary
Some embodiments of the present invention pertain to an apparatus, method, and a computer program that is configured to cause a processor to construct an upper envelope and lower envelope for an input image based on a statistically computed window parameter in a real-time system using multi-thread processing, and smooth the upper envelope and lower envelope based on a sum of elements in a window, wherein the sum of elements in the window is divided by a number of non-zero elements in the sum of elements in the window.