Colored drawing cultural relic imaging hyperspectral nondestructive detection and recognition system under deep learning

Abstract

The invention provides a colored drawing cultural relic imaging hyperspectral nondestructive detection and recognition system under deep learning. The colored drawing cultural relic imaging hyperspectral nondestructive detection and recognition system comprises a lens and filtering subsystem, a high-speed imaging subsystem and an analysis processing module, wherein the lens and filtering subsystemis used for collecting hyperspectral images of cultural relics under the control of the high-speed imaging subsystem, the high-speed imaging subsystem is used for controlling working parameters of the lens and filtering subsystem and realizing transmission and storage of hyperspectral images, and the analysis processing module is used for processing and analyzing the acquired hyperspectral images, completing identification and information extraction of cultural relics and displaying results in a graphical interface mode. The system is advantaged in that rapid identification and information acquisition of the colored drawing cultural relics can be carried out, a professional, imaging, rapid measurement and intelligent cultural relic information acquisition detection and identification system is realized, and requirements of cultural relic protection and research field on a cultural relic special spectrometer are met.

Description

technical field [0001] The invention relates to the technical field of target classification and recognition, in particular to a hyperspectral nondestructive detection and recognition system for painted cultural relics imaging under deep learning. Background technique [0002] Hyperspectral imaging technology is mostly used in the identification and information mining of painted cultural relics such as murals, ancient paintings, oil paintings, and ancient manuscripts. At present, there are few special instruments in the field of highly reproducible nondestructive testing and identification of painted cultural relics. [0003] For painted cultural relics, existing spectrometers have the following problems: 1) The imaging band range is narrow and cannot completely cover the bands required for cultural relics research; 2) There is no effective band selection, and the effective spectral measurement range of painted cultural relics is not determined. There are many useless bands ...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to provide a hyperspectral non-destructive detection and identification system for painted cultural relics imaging under deep learning, so as to at least partially solve the problems of large volume, slow imaging speed and inability to target painted cultural relics in existing imaging spectroscopy instruments. On-site rapid identification of cultural relics

Images