| In the past decade colorimetry has been rapidly developed.It is key technology in many industries, such as colourte-levision, colour imaging, dyeing, and paint and so on. Colourcan be described by its tristimulus values, and many kinds oftristimulus values has been used in the colour industry, suchas RGB, which are used in the computer and some other devices,and XYZ, which are prescribed by CIE and is a independent colourspace. The tristimulus values XYZ can be transformed to otherspace, such as RGB, CMYK, Lab, and etc., and be treated as abridge of colour industry. So learning method on computingtristimulus values and compare precision of these method arevery necessary in the industry of colour.Using of weighting tables makes the computation of CIEtristimulus values simpler and more conveniently. So theprecision of computation of tristimulus values lies mostly onthe characteristics of weights. The CIE has standardized themethod for computing tristimulus values. But CIE has nevergiven any guidance as how to prepare the weights, hencesignificant differences can be obtained for computing thetristimulus values of the same color object when differentweighting tables are used. In order to overcome this problem,American Society for Testing and Materials (ASTM) had publishedtwo kinds of standard weighting tables known as ASTM Table5and Table6. However, there still are many issues needed to befurther investigated, such as the ASTM standard weightingtables can be improved or not? Besides, the ASTM standardweighting tables are only for standard light sources. However,in practice the light sources used are different from the standard ones, hence the user have to prepare their ownweighting tables. But there is little work existed. The mainobjective of this work is to study the existed methods forcomputing weighting tables and finally make comprehensivecomparisons among the existed and newly developed methods.In the part of computing weighting tables, in order toprovide weighting tables required by industrial applications,and provide scientific evidence for CIE and ASTM to standardizeweighting tables and method for computing weighting tables, wefirstly study the CIE standard method for computing tristimulusvalues, and the ASTM standard weighting tables and method forcomputing weighting tables. Hence an explicit and simplifiedmethod for computing ASTM weighting tables of Table5is given.Secondly, in this thesis, a new method based on Spragueinterpolation is given for computing wighting tables. Thirdly,Oleari method based on deconvolution is studied and implemented.Besides, a new method based on Oleari’s work is given forcomputing weighting tables. Fourthly, a optimum method and aleast squares method are given for generating weighting tables，and introduced the method on computing weighting tables inindustry applications and so on. Fifthly, methods for computingweighting tables for a shorter range of wavelength interval areintroduced. Finally, most methods are numerically compared andconclusions are given. |
