Development And Application Of A Multi Type Hyperspectral Imaging Integrated System

Spectral imaging technology obtains the image of the observation object in a specific band by wave filtering, which has high spectral and spatial resolving power. Spectral imaging technology not only can be used to detect objects in qualitative and quantitative way, but also be used to locate the targets. Therefore, spectral imaging technology has been widely used in various research areas. Different kinds of hyperspectral imaging system can obtain different types of spectral images, such as absorption spectra, reflectance spectra, fluorescence spectra and so on. For a single spectral imaging system, its applications are limited by the sample. Additionally, each single spectral imaging system has its advantages and limitations. At present, there is no spectral imaging technology can be applied to the detection of all kinds of sample. In order to make full use of various spectral imaging techniques, a hyperspectral imaging system integrated with multi type spectral detection capability is developed in this paper. This multi type hyperspectral imaging system can extend the application of spectral imaging technology.According to the analysis of the feasibility of the system, a multi type spectrum(transmission spectrum, reflectance spectrum and fluorescence spectrum) imaging integrated system with a liquid crystal tunable filter(LCTF) were developed. At the same time, the performance of this system has been investigated. The synchronous control software of this system enables the LCTF and CCD camera to work together and obtains the spectral image data. And the spectral image data analysis software of this system are used to analyze the acquired image spectral data.The spectral image data analysis software of this system provides region of interest(ROI) sampling tools for the extraction of a single point, free curve, polygon, rectangle, etc. The method which classify the pixels in accordance with the classification decision rule, make the system having a strong spectral matching classification function for the sample. The ability of spectral mixture decomposition makes the distinguishing of the interest target from other noninterest content more easily, and greatly improves the efficiency of the sample testing.By using the system developed in this paper, the feasibility of using hyperspectral imaging technology in detecting the existence of illicit poppy cultivation is studied. Experimental results show that poppy leaf can be identified by the spectral matching classification function of hyperspectral imaging technology, and the difference of the leaves in the surrounding environment can be distinguished. This solves the qualitative problems of opium poppy cultivation. The quantitative problem of the poppy cultivation area can be solved by using spectral classification and the proportion of poppy cultivation in the target can be estimated statistical techniques. The target signal of the poppy can be extracted and the processing results can be output in pseudo color display mode with the spectral decomposition technique, which solves the problem of poppy target localization in complex environment. Experiments confirm that it is feasible to examine the existence of the poppy cultivation area with spectral imaging method.In addition to the non-contact, non-damage, high accuracy, good repeatability and many other advantages, the multi type hyperspectral imaging integrated system developed in this paper, also has the advantages of compact structure, easy operation, portability, and high spectral image data processing ability. It has positive significance in the popularization and application of hyperspectral imaging technology.