Hyperspectral acquisition method, hyperspectral acquisition system, electronic equipment and coded wide-spectrum imaging device

Abstract

The invention discloses a hyperspectral acquisition method, a hyperspectral acquisition system, electronic equipment and a coding wide-spectrum imaging device. The hyperspectrum acquisition method comprises the following steps: acquiring brightness data of a broadband spectrum light source and brightness data of background light; preprocessing the brightness data of the broadband spectrum light source according to the brightness data of the background light to obtain the brightness data of the broadband spectrum light source without the background light; performing similarity classification processing on the broadband spectrum light source brightness data with the background light removed according to a similarity classification model so that a classification lookup table of the broadband spectrum light source brightness data is obtained; and resolving and restoring the broadband spectrum light source data without the background light into spectrum curve data of the measured object according to the classification lookup table of the broadband spectrum light source brightness data so as to complete the acquisition of the hyperspectrum.

Description

Technical field [0001] The present invention relates to the field of high spectroscopy imaging, and more particularly to high spectral acquisition methods and systems, electronic devices, and encoded wide spectra imaging devices. Background technique [0002] The spectral imaging technology uses the absorption or radiation characteristics of the substance to different electromagnetic spectroscopy. On the basis of normal two-dimensional space imaging, it can increase one-dimensional spectral information, which can be used in ultraviolet, visible light, near infrared and medium in electromagnetic band. In the infrared area, many narrowbands and spectral continuous image data are obtained, providing a complete and continuous spectral curve for each cell. Depending on the band of the target obtained, the spectral imaging technique can be divided into multi-spectral imaging techniques (acquisition bands between 3-12, the spectral resolution is generally 100 nm), high spectroscopy tech...

AI Technical Summary

Problems solved by technology

[0004] However, on the one hand, the disadvantages of spectral imaging devices equipped with grating spectroscopy, acousto-optic tunable filter spectroscopy, prism spectroscopy, or chip coatings are mainly that the spectral resolution is limited by the spectroscopic accuracy of the gratings, filters, prisms, or coatings. The cost of realizing hyperspectral imaging is high, and at the same time, after the spectrum is narrow-band splitted, the energy of the narrow-band light obtained is greatly reduced, which will also affect the imaging quality; on the other hand, the disadvantage of the spectral imaging device equipped with encoded broadband imaging is that each imaging The hyperspectral restoration of points requires a large amount of calculation

In particular, considering that the resolution of commonly used image sensors is over one million or even tens of millions of pixels, the computational load for spectral restoration of all pixels is very large, which will lead to encoding wide-spectral imaging methods on general embedded low-power platforms. Difficult to implement

Images