The Spectral Reflectance Reconstruction Based On A Digital Camera

According to the needs of different people,we are trying to show the color we saw through various methods.However,when compare with the original colors together,we often find that there is more or less difference between both.Because of the Same Color Spectrum Phenomenon,we can't realize color copy really,it is difficult to meet the requirements of color reiteration for high quality and high precision.We can convert ordinary image to spectral image,obtained by the spectral reflection ratios to achieve accurate color reproduction.This paper mainly studies the reconstruction of the spectral reflectance scheme based on digital camera.Using a digital camera to get the target surface color of RGB signals.Obtain the object color sample data and the digital camera color information preprocessing by studying how to use a digital camera,obtain more accurate spectral reflectance based on digital camera samples.Through the analysis of research status at home and abroad,raises the main research content.Through the analysis of color matching standard,raises the spectrum matching more accurately through the representation of the color on the surface of the object.Through analyzing the phenomenon of same color spectrum influence on color accurate emersion,put forward the basis of spectral reflectance reconstruction.The digital camera color information preprocessing and linear correction on digital response values through the elaboration of color sample information acquisition.Enumerate the accuracy of spectral reflectance evaluation method and the pre-processing results were analyzed,and one of its innovative is using linear calibration,the calibration of camera digital value,correction before the Root Mean Square Error,Color Difference Method,Peak Signal-to-Noise Ratio Method,Coefficient of Goodness-of-Fit are:0.0632,5.9831,86.7813,0.9836.After correction,the results are:0.0514,5.6083,88.3624,0.9978.Illustrates the necessity of the pretreatment of the color information of digital cameras.In this paper,another innovative point is that the three main reconstruction algorithm of spectral reflectance was improved.By introducing Simpson formula and Lagrange interpolation function to carry on the smooth constraint solving,the Smooth Constraint Condition was improved.Through the Pseudo-Inverse Method from the Angle of mathematics the points of discussion,introduces the least norm and Lagrange multiplier method,the method of solving the pseudo-inverse method is improved.Through the correlation matrix C=S·S~T(S is the group of spectral reflectance samples),the Principal Component Analysis(PCA)method was improved.Respectively for three kinds of reconstruction accuracy has made the detailed analysis and evaluation,illustrates the significance of spectral reflectance reconstruction.Finally in this paper,research work was summarized,on the basis of work on subsequent further research prospects.