The Color Measurement Of The Spectrum And Recurrent Form Monitor Based On The Digital Camera

The rise of color science has provided great help for domains like image data processing anddyeing businesses, remote transmission of e-commerce pictures, films and television programscommunication, telemedicine, and artworks collection, etc. while bringing about convenience forpeople’s routine work and life. With the arrival of the digital informatization age, the abovedomains have higher and higher demand for color quantification and display recurrence. It hasalready become a fact that methods with RGB value, chromaticity coordinate value and CMYKvalue to represent colors can hardly fulfill the demand in terms of accuracy; whereas thecross-device color transmission method based on RGB value and CIE chromaticity coordinateconversion has appeared complicated and inefficient. Therefore, research on color quantificationand cross-device transmission on the spectral scale seems extremely crucial and urgent.With the RGB values of samples as the initial data, this paper acquires and corrects RGB datavia a digital camera, and defines the joint space of spectrum colors to map the RGB values ofsamples. A prediction is made from RGB value to spectral reflectance through polynomialmodeling and principal component analysis (PCA). Finally, the method of simulating CIEchromaticity coordinate values with spectral reflectance is adopted for cross-device displayrecurrence of colors. This paper mainly conducts the following studies:(1) Acquisition of RGB signal and correction based on camera characteristics and light source:This paper proposes the idea of using digital camera to acquire the RGB values of samples on thebase of summarizing and comparing the traits and operation modes of spectrophotometer, scannerand digital camera, and performs spatial and linear correction over the RGB values on the basis offactors like arranged light source in the shot environment and camera’s imaging properties, whichreduces complexity for spectral reconstruction.(2) By analyzing and summarizing several joint spaces of spectrum colors (principalcomponent space, Lab space, Lab combined principal component space, and Lab combinedchromaticity space), this paper expounds their limitations, on this base puts forward the joint spaceof spectrum colors in combination with CIE stimulus value and human-eye stimulus, and uses duallight sources to modify the defined space while using root-mean-square verification to improve thevalidity of the space.(3) This paper conducts analysis and summary of the current spectral reconstruction methodsand extends the commonly used one (PCA). On the base of combining the polynomial model, itbuilds the spectral reconstruction model by making an LUT color card as sample which contains all color lumps ranging from white to black.(4) By analyzing the complexity and limitation of the method of cross-device colortransmission from RGB values to CIE chromaticity coordinate values, it raises the implementationof image display recurrence through reverse chromaticity characterization over the display with thereconstructed spectral reflectance to simulate the CIE chromaticity coordinate values. Withmultispectral images as samples, EIZO CG241LCD as display device, and Canon D70digitalcamera as input device, a series of operations from spectral reconstruction to display recurrence areperformed to verify the validity of the method raised and the model built.