摘要: |
目的 综述目前已有光谱反射率重建技术的研究进展,为光谱反射率重建技术的适用范围更广和精度更高提供科学的研究基础。方法 介绍目前已有光谱反射率重建技术的相关方法,并探讨相关方法的优缺点和适用的应用场景,提出相关方法的改进以及未来发展的趋势。结果 直接重建法的重建精度相对较低,插值重建只适用于小样本的数据集,基于人工智能技术虽然精度高,在实际应用中也受到局限,组合重建法将各种重建方法按照一定的方式组合起来,并加入相关的修正项进行光谱重建,解决了其他3类方法存在的缺陷,保证了光谱反射率重建精度,解决了真正实际应用需要。结论 通过组合的形式解决了在重建光谱反射率过程中遇到的噪声、智能化等问题,通过组合的方法可以解决各自单一方法缺陷,进行互补。 |
关键词: 直接重建法 插值重建法 基于人工智能的重建法 组合重建法 |
DOI:10.19554/j.cnki.1001-3563.2020.11.037 |
分类号:TS801.9 |
基金项目:上海市自然科学基金(19ZR1435900) |
|
Research Progress of Spectral Reflectance Reconstruction Technology |
WANG Wen-ju, WANG Jiang-wei
|
School of Publishing, Printing and Art Design, University of Shanghai for Science and Technology, Shanghai 200093, China
|
Abstract: |
The work aims to summarize the current research progress of the spectral reflectance reconstruction tech-nology, so as to provide a scientific research foundation for the broader application and higher accuracy of the spectral reflectance reconstruction technology. The related methods of the existing spectral reflectance reconstruction technology were introduced, the advantages and disadvantages of the related methods as well as applicable application scenarios were discussed and the improvement of the related methods and the future development trends were proposed. The reconstruction accuracy of the direct reconstruction method was relatively low and the interpolation reconstruction was only suitable for small sample data sets. Although the accuracy based on artificial intelligence technology was high, it was also limited in practical applications. Therefore, the combination reconstruction method was recommended to combine various recon-struction methods in a certain way and related correction items were added to perform spectral reconstruction, thus solving the defects of the other three methods, ensuring the accuracy of spectral reflectance reconstruction, and meeting the needs of real practical applications. The combination form solves the problems of noise and intelligence encountered in the process of reconstructing the spectral reflectance and the combination method can solve the defects of each single method by complementing each other. |
Key words: direct reconstruction interpolation reconstruction reconstruction method based on artificial intelligence combined reconstruction |