Spectrum Data Correction via Adaptive Moving Average
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional spectrum data correction methods risk applying noise removal to signal peaks and can cause data discontinuity, especially in spectra with abrupt signal level changes, and fail to accurately distinguish between noise and signal, leading to potential misapplication of noise removal processes.
Innovation Solution
A spectrum data correction system and method that detect noise based on the difference between extremal points and the mean value of nearby data blocks, controlling the count of moving-average points according to the noise quantity and signal level, ensuring accurate noise removal without affecting signal peaks and reducing data discontinuity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If moving average processing is applied to all measurement data-blocks, then noise removal is achieved, but spectral shape rounding occurs
Solution Approach 1:
The patent applies moving average processing selectively only to measurement data-blocks identified as containing noise, rather than uniformly to all data-blocks. The noise detection mechanism identifies specific regions where noise is present, and the moving average filter is applied locally to those regions only, preserving the original spectral shape in noise-free regions while removing noise in affected regions.
2Shape
If user selects measurement data-blocks for localized moving average processing, then spectral shape rounding is prevented, but automation is lost
Solution Approach 1:
The system automatically detects noise in measurement data-blocks using a noise detection mechanism that compares signal characteristics against noise thresholds. This self-service approach eliminates the need for manual user selection of data-blocks requiring noise removal, while still applying localized moving average processing only where needed, thus maintaining both spectral shape integrity and full automation.
3Object-affected harmful factors
If noise removal is applied to steep spectral peaks, then noise is removed, but signal distortion occurs
Solution Approach 1:
The noise detection mechanism first identifies whether a measurement data-block contains noise before applying moving average processing. By detecting the presence of noise in advance and only applying the filter when noise is confirmed, the system prevents unnecessary filtering of steep spectral peaks, thus avoiding signal distortion while still removing noise where present.
Data Source
AI summary
An object of the invention is to automatically remove noise out of an input spectrum data while rounding in spectral shape is being suppressed. The spectrum data correction system wherein a noise quantity is detected on the basis of a difference in value between an extremal point of measurement data-blocks making up input spectrum data, and a mean value of measurement data-blocks in the vicinity of the extremal point, and a count of moving-average points of measurement data-blocks subjected to moving average processing is controlled according to the noise quantity, the spectrum data correction system comprises means for using an optical power, detected by a photo-detector of an optical spectrum analyzer, as the measurement data-block, thereby controlling the count of the moving-average points of the measurement data-blocks according to magnitude of a signal level of the measurement data-block.