Study On Spectral Reflectance Reconstruction Of Biological Tissue Images

Multispectral and hyperspectral imaging technology has become a popular application in the field of biomedicine,providing more reliable technical support for medical imaging and pathological analysis.However,most of the current multispectral and hyperspectral imaging in the biomedical field have problems such as expensive equipment,low efficiency and poor real-time application.The research on the reconstruction of spectral information from the color images of biological tissues is a hot research field at present,which has important theoretical value and application prospect for improving the level of medical imaging and analysis.In this paper,the color image and spectral reflectance information of biological tissue are taken as the research object,focusing on the field of spectral reflectance reconstruction based on color image,and a spectral reflectance reconstruction method suitable for biological tissue image is proposed.The main research work is as follows:(1)In view of the spectral color characteristics of biological tissues and the limitations of current spectral reflectance reconstruction methods,this study proposed a new spectral reflectance reconstruction algorithm based on lattice regression and metameric black,which can maintain high spectral reconstruction accuracy and efficiency while maintaining the consistency of the physical layer in the reconstruction process,and has a high color restoration accuracy(2)Aiming at the problem that there are few public spectral data sets in the biomedical field,this study uses a multi-spectral imaging system to capture spectral reflectance images of multiple groups of biological tissues,which provides a strong data support for this research topic.(3)For the proposed spectral reflectance reconstruction algorithm,two sets of experiments were set up in restricted environment and open environment to fully verify the effectiveness of the algorithm.Firstly,the data acquisition process and experimental process were designed in a restricted environment to verify the spectral reflectance reconstruction accuracy of the proposed algorithm for simulated biological tissue color images.Secondly,different data acquisition and experimental processes are designed in an open environment to verify the spectral reflectance reconstruction effect of the proposed algorithm on real biological tissue color images.The two sets of experiments jointly verified the effectiveness of the proposed method and technical path for spectral reflectance reconstruction of biological tissue images.