Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Model transferring method based on dual-tree complex wavelet and piecewise direct standardization

A dual-tree complex wavelet and model transfer technology, which is applied in special data processing applications, instruments, electrical digital data processing, etc., can solve the problems of insufficient precision, general model transfer, poor translational variability, etc., and achieve the effect of wide application

Inactive Publication Date: 2016-06-01
河北伊诺光学科技股份有限公司
View PDF3 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The segmented direct normalization algorithm corrects the entire spectrum and does not divide the spectrum in the frequency domain. Such a model transfer is very general and not fine enough.
However, the wavelet multi-scale segmental direct normalization algorithm based on the segmental direct normalization algorithm has the characteristics of multi-scale correction to overcome the general and not precise shortcomings of the segmental direct normalization method, but the translational variability of wavelet has caused This method is poor at correcting for drift in the x-axis of the spectra of the two instruments

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Model transferring method based on dual-tree complex wavelet and piecewise direct standardization
  • Model transferring method based on dual-tree complex wavelet and piecewise direct standardization
  • Model transferring method based on dual-tree complex wavelet and piecewise direct standardization

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0060] (1) Select the corn sample as the sample to be tested, use the chemical method to measure the humidity data of the corn sample, and then collect the spectral data of the master and slave instruments of the sample; collecting spectral data is to collect the spectral data of the same sample under the same measurement conditions. Master and slave refer to two different instruments of the same model.

[0061] As shown in Figure 1(a) and Figure 1(b), the master and slave instruments in this example are two near-infrared spectroscopy instruments m5 and mp5, m5 is the master and mp5 is the slave, as figure 2 As shown, there is a difference between the m5 instrument spectrum and the mp5 instrument spectrum.

[0062] (2) Use the Kennard-Stone algorithm to optimize the corn sample spectrum, remove the abnormal sample spectrum in the sample spectrum, and then divide the sample spectrum into a training set spectrum and a test set spectrum.

[0063] (3) Using the optimal decomposit...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a model transferring method based on a dual-tree complex wavelet and piecewise direct standardization. The method comprises the steps of firstly adopting the dual-tree complex wavelet to perform multi-scale decomposition and reconstruction, then correcting each layer of reconstructed spectrum by using the piecewise direct standardization, then establishing prediction models based on cross validation of a partial least squares method and a leave-one-out method, and at last fusing all prediction models by computed weights and evaluating. According to the method combining the dual-tree complex wavelet with the piecewise direct standardization provided by the invention, as the dual-tree complex wavelet has the characteristics of translation invariance and multi-scale, the deficiencies of the existing model transferring method are overcome, the capability of correcting drifting in X-axis and Y-axis directions is very excellent, and meanwhile the advantages of being fine, accurate and efficient are provided, and thereby the method can be widely applied to the fields of near-infrared and Raman spectrum.

Description

technical field [0001] The invention relates to a model transfer method, in particular to a model transfer method based on double-tree complex wavelet segmentation direct standardization. Background technique [0002] Model transfer is also called instrument standardization, which means that after mathematical processing, the model on one instrument can be used for another instrument, thereby reducing the huge workload caused by remodeling and realizing the sharing of samples and data resources. At present, the most widely used and most successful method is the multivariate correction algorithm, and among them, the piecewise direct standardization (PDS) algorithm and the improved algorithm based on it have better effects. [0003] For the spectrum, the difference between the spectra of the two instruments can be summarized into two aspects, one is the difference in the x-axis direction, that is, the shift of the waveform; the other is the difference in the y-axis direction, ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F19/00
Inventor 陈达卢帆
Owner 河北伊诺光学科技股份有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products