Hyperspectral Dimensionality Reduction Matching Method and System Based on Spectral Sampling Histogram
A histogram and hyperspectral technology, applied in the field of hyperspectral dimensionality reduction matching, can solve problems such as poor performance, achieve the effects of good real-time performance, reduce the amount of matching operations, and improve accuracy
Inactive Publication Date: 2017-02-15
WUHAN UNIV
View PDF2 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, judging from the research status, there is no dimensionality reduction matching algorithm specifically for hyperspectral data. The current spectral dimensionality reduction algorithms are all designed for hyperspectral data with hundreds of dimensions. If they are directly used for dimensionality reduction of hyperspectral data and then match, its poor performance
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0032] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0033] Refer to attached figure 1 , the present invention mainly consists of four steps: spectral normalization, obtaining a sampling histogram of the spectrum, calculating the Euclidean distance, and taking the minimum Euclidean distance to complete the matching. Example A spectral library containing 1432 substance spectra is selected, and each substance has only one spectral data in the spectral library. The spectral resolution is Δσ=0.1cm -1 , the wavelength range is 2-14μm, and the corresponding wavenumber range is 5000-714cm -1 . There are N=42861 sampling points in this spectral range, that is, the original spectrum has 42861 dimensions.
[0034] During specific implementation, the technical solution of the present invention can use computer software technology to realize the automatic operation process. The implementation steps of the embodi...
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
Login to View More Abstract
The invention provides a hyper spectrum dimensionality reduction matching method and system based on spectrum sampling histogram. The hyper spectrum dimensionality reduction matching method comprises the following steps: respectively performing normalization processing on a spectrum to be matched and all spectrums in a spectrum bank, respectively acquiring the sampling histogram of the normalized spectrum to be matched and the sampling histogram of all the normalized spectrums in the spectrum bank, calculating the Euclidean distance of the sampling histogram of the spectrum to be matched and the sampling histogram of all the spectrums in the spectrum bank, and selecting the spectrum with the smallest Euclidean distance to the sampling histogram of the spectrum to be matched from the spectrum bank as a matching target. According to the hyper spectrum dimensionality reduction matching method, as the normalized spectrums are sampled by using barrow bands with equal intervals, the sampling histogram of which the dimensionality is much smaller than that of an original spectrum is obtained, the dimensionality reduction of the spectrum is completed, the sampling histogram subjected to dimensionality reduction is adopted to replace the original spectrum for matching, the calculation amount in follow-up matching is greatly reduced, the relative position information in the spectrum is maintained by using a sectional extraction method in sampling, and the precision of matching is improved.
Description
technical field [0001] The invention relates to the technical field of hyperspectral dimensionality reduction matching, in particular, the invention relates to a hyperspectral dimensionality reduction matching method and system based on a spectral sampling histogram. Background technique [0002] Spectral remote sensing instruments are developing toward higher spectral resolution, spatial resolution, and time resolution, and the dimensions of the measured spectral data can reach thousands or even tens of thousands of dimensions. After obtaining hyperspectral data, the most common and basic requirement is to identify the type of substance, and identification of the type requires spectral matching. Simply put, spectral matching is to compare all the reference spectra in the spectral database with the spectrum to be measured one by one by calculation, and select the most similar spectrum as the matching result of the spectrum to be measured. The matching results can be used to...
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
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06T7/00G06T7/30
CPCG06T3/0031G06T7/30
Inventor 黄珺马佳义
Owner WUHAN UNIV