US 10,359,404 B2Granted

Noise level estimation method, measurement data processing device and program for processing measurement data

AssigneeShimadzu Corporation

InventorHiroaki Kozawa

Life of this patent

Estimating the term…

Loading patent…

Abstract

In a method for estimating a noise level representing the magnitude of a noise component from measurement data, first waveform data composed of high frequency noise components extracted from assumed noise data are divided into segments so that each section where positive values successively occur or each section where negative values successively occur in the first waveform data is defined as one segment. A segment-width threshold is determined based on the distribution of the widths of the segments. Second waveform data composed of high frequency noise components extracted from measurement data are divided into segments in the same manner. Each segment having a width larger than the threshold is excluded from the segments in the second waveform data, to create a first segment group. The noise level is determined based on the heights or areas of the plurality of segments included in the first segment group.

Claims12

Dependent claims

A noise level estimation method for estimating a noise level representing a magnitude of a noise component from measurement data containing three components of a peak component, baseline component and noise component, the method comprising:

a) extracting, from assumed noise data purely composed of assumed noise for the measurement data, high frequency noise components by means of a high-pass filter which reduces frequency components lower than a predetermined frequency, dividing first waveform data which are waveform data of the high frequency noise components into a plurality of segments so that each section where positive values successively occur or each section where negative values successively occur in the first waveform data is defined as one segment, or so that each section between a local maximum and a local minimum neighboring each other in the first waveform data is defined as one segment, and determining a segment-width threshold based on a distribution of widths of the plurality of segments;
b) extracting high frequency measurement components from the measurement data by means of the high-pass filter, and dividing second waveform data which are waveform data of the high frequency measurement components into a plurality of segments so that each section where positive values successively occur or each section where negative values successively occur in the second waveform data is defined as one segment, or so that each section between a local maximum and a local minimum neighboring each other in the second waveform data is defined as one segment;
c) excluding each segment having a width larger than the threshold from the plurality of segments obtained by dividing the second waveform data, to create a first segment group formed by the remaining segments;
and
d) determining the noise level based on heights or areas of the plurality of segments included in the first segment group.

2 more elements

A measurement data processing device used for estimating a noise level from measurement data containing three components of a peak component, baseline component and noise component, the device comprising:

a) a high-pass filter for reducing frequency components lower than a predetermined frequency;
b) a storage for holding a segment-width threshold, the threshold previously determined by extracting, from assumed noise data purely composed of assumed noise for the measurement data, high frequency noise components by means of the high-pass filter, dividing first waveform data which are waveform data of the high frequency noise components into a plurality of segments so that each section where positive values successively occur or each section where negative values successively occur in the first waveform data is defined as one segment, or so that each section between a local maximum and a local minimum neighboring each other in the first waveform data is defined as one segment, and determining the threshold based on a distribution of widths of the plurality of segments;
and
a processor configured to
extract high frequency measurement components from the measurement data by means of the high-pass filter, and for acquiring second waveform data which are waveform data of the high frequency measurement components;
divide the second waveform data into a plurality of segments so that each section where positive values successively occur or each section where negative values successively occur in the second waveform data is defined as one segment, or so that each section between a local maximum and a local minimum neighboring each other in the second waveform data is defined as one segment;
exclude each segment having a width larger than the threshold from the plurality of segments obtained by dividing the second waveform data, to create a first segment group;
and
determine the noise level based on heights or areas of the plurality of segments included in the first segment group.

6 more elements

Dependencies are inferred from the claim text.

Facts

Application: 15/544,036Filed 26 Jan 2015
Published: US 2018/0011067 A111 Jan 2018
Granted: 23 Jul 2019
Term ends: Not estimatedThe source records don’t support an estimate.

Classification5 codes

G01N 30/861
With integration or differentiation: differentiation
G01N 30/8641
Detection of slopes or peaks; baseline correction: baseline
G01N 30/8675
Signal analysis: evaluation, i.e. decoding of the signal into analytical information
G01N 30/8686
Evaluation, i.e. decoding of the signal into analytical information: fingerprinting, e.g. without prior knowledge of the sample components