光谱成像技术研究室知识库

With the development of spectral imaging towards higher space resolution, higher spectral resolution and higher signal to noise ratio, some problems have appeared in the traditional spectral imager, for example, data acquisition quantity is too big, the resolution is affected by frame frequency and pixel size of detector, precise alignment is difficult for big caliber and long focus system, and hard to develop signal to noise ratio because of limited optics power. To solve the above problems, a single dispersion spectral imager based on compressed coding is studied. Specially, for the lack of system realization and experiment verification at home, the designation, realization, mathematic model and reconstruction algorithm under multi-frame measurement are mainly studied, and the prototype testing and data processing are achieved. At last, some key problems still need to study, such as code error analysis, multi-model and multi-algorithm, system demarcation, and reconstruction evaluation. This imaging system is consisted of object glass, coding template, dispersion element, collimating lens, focus lens and detector, and hyperspectral data was reconstructed by sparse reconstruction algorithm. There are many advantages in the new system, for example, a smaller data size due to the sparse sample of multi-information, a higher resolution because of code super-resolution, an easier implementation for lower hardware requirement, a higher optical energy usage because the code is instead of slit. The results show that the measurement is efficient, the design of prototype is proper, reconstruction algorithm and calibration method are accurate, the space infoimation of alphabet HSI object is clear, and the spectral information of alphabet HSI object is accurate and closed to tungsten lamp spectral, so the system designation and engineering realization are feasible.