A study of digital camera colorimetric characterization based on polynomial modeling

The digital camera is a powerful tool to capture images for use in image processing and colour communication. However, the RGB signals generated by a digital camera are device‐dependent, i.e., different digital cameras produce different RGB responses for the same scene. Furthermore, they are not colorimetric, i.e., the output RGB signals do not directly correspond to the device‐independent tristimulus values based on the CIE standard colorimetric observer. One approach for deriving a colorimetric mapping between camera RGB signals and CIE tristimulus values uses polynomial modeling and is described here. The least squares fitting technique was used to derive the coefficients of 3 × n polynomial transfer matrices, yielding a modeling accuracy typically averaging 1 ΔE units in CMC(1:1) when a 3 × 11 matrix is used. Experiments were carried out to investigate the repeatability of the digitizing system, characterization performance when different polynomials were used, modeling accuracy when 8‐bit and 12‐bit RGB data were used for characterization, and the number of reference samples needed to achieve a reasonable degree of modeling accuracy. Choice of characterization target and media and their effect on metamerism have been examined. It is demonstrated that a model is dependent upon both media and colorant, and applying a model to other media/colorants can lead to serious eye–camera metamerism problems. © 2001 John Wiley & Sons, Inc. Col Res Appl, 26, 76–84, 2001     

Concerning the calculation of the color gamut in a digital camera

Several methods to determine the color gamut of any digital camera are shown. Since an input device is additive, its color triangle was obtained from their spectral sensitivities, and it was compared with the theoretical sensors of Ives‐Abney‐Yule and MacAdam. On the other hand, the RGB digital data of the optimal or MacAdam colors were simulated to transform them into XYZ data according to the colorimetric profile of the digital camera. From this, the MacAdam limits associated to the digital camera are compared with the corresponding ones of the CIE‐1931 XYZ standard observer, resulting that our color device has much smaller MacAdam loci than those of the colorimetric standard observer. Taking this into account, we have estimated the reduction of discernible colors by the digital camera applying a chromatic discrimination model and a packing algorithm to obtain color discrimination ellipses. Calculating the relative decrement of distinguishable colors by the digital camera in comparison with the colorimetric standard observer at different luminance factors of the optimal colors, we have found that the camera distinguishes considerably fewer very dark than very light ones, but relatively much more colors with middle lightness (Y between 40 and 70, or L* between 69.5 and 87.0). This behavior is due to the short dynamic range of the digital camera response. © 2006 Wiley Periodicals, Inc. Col Res Appl, 31, 399–410, 2006; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/col.20245     

A strategy toward spectral and colorimetric color reproduction using ordinary digital cameras

In this work, a methodology is introduced to use ordinary digital RGB cameras for the purpose of spectral and colorimetric color reproduction. First, it is attempted to recover the spectral reflectance from RGB camera response using different approaches, among which, it is shown that weighted nonlinear regression as performed better than other approaches. After analyzing the results, it is realized that although spectrally the results are satisfactory, there is still a significant colorimetric error left, that should be addressed. Therefore, in the second part of the article, different linear and nonlinear matrix transforms are used to change the RGB camera response to CIEXYZ tristimulus values under a specific condition. It is concluded that colorimetric error of the recovery can be reduced significantly when a separate path is used for colorimetric color reproduction.     

Colour‐imager characterization by parametric fitting of sensor responses

This study describes a novel method for characterizing the colorimetric and photometric properties of three‐channel color imaging devices. The method is designed to overcome some undocumented aspects of the imager‐characterization problem: The effective spectral sensitivity profiles of the imager's color channels depend on the level of radiant input energy, and these profiles must be known in order to determine the true intensity‐response characteristics of the three channels. By fitting the response distributions of the three color channels explicitly with low‐dimensional models, the method takes these dependencies into account, and may, therefore, offer several advantages over other imager‐characterization methodologies, particularly where illuminant‐independent characterization is required. An application of the technique is detailed, in which a CCD camera is characterized using only the Macbeth ColorChecker and a number of artificial illuminants. © 2001 John Wiley & Sons, Inc. Col Res Appl, 26, 442–449, 2001     

An efficient nonlinear polynomial color characterization method based on interrelations of color spaces

A well-known color characterization method is to take an image of a color chart and then to find the mapping matrix from the digital RGBs to the corresponding known CIEXYZs. However, the prediction errors are generally large in CIELAB color space because of the nonlinear transformation from CIEXYZs to CIELABs. In this article, we propose an efficient and simple nonlinear method for the color characterization of input devices. The approach for deriving a colorimetric mapping between digital RGB signals and CIELAB tristimulus values uses the polynomial modeling by considering the interrelations among the standard CIE color spaces. Furthermore, to improve the accuracy of solution, we take the polynomial root terms extension. Our algorithm is simple to implement because only a least-squares mapping should be solved. Various computational results are given to demonstrate the efficiency and capability of the proposed method.     

Estimation of optoelectronic conversion functions of imaging devices without using gray samples

In the colorimetric or spectral characterization of imaging devices such as digital cameras and scanners, the optoelectronic conversion functions (OECFs) are traditionally obtained from standard gray samples. However, these gray samples are sometimes unavailable when conducting color characterization. We propose an efficient method for recovering OECFs by using nongray samples, based on the finite‐dimensional modeling of spectral reflectance and the second‐order polynomial fitting of OECFs. Experimental results indicate that the accuracy of the estimated OECFs are close to those obtained from gray samples, with the correlation coefficients R² larger than 0.995. The proposed method should be useful in colorimetric and spectral characterization of imaging devices by using custom‐made color samples in textile or other industries, when standard gray samples are not available and not easily made. © 2008 Wiley Periodicals, Inc. Col Res Appl, 33, 135–141, 2008     

Digital color reconstructions of cultural heritage using color‐managed imaging and small‐aperture spectrophotometry

There are many examples of cultural heritage having optical properties that have changed with the passage of time. Examples include the yellowing, darkening, and fading of paints and varnishes caused by light exposure and atmospheric pollution. When it is infeasible to treat an object, an image simulation can provide a view to the past, known as a color reconstruction. A technique is described that relies on a color‐managed image, spectral reflectance factor measurements of the object, an optical model of colorant mixing, an optical database of artist materials, spreadsheet software, and image editing software. Spectral calculations are used to create adjustment curves where segmented portions of an object's image are translated in color. This approach has been used to produce color reconstructions of paintings by Vincent van Gogh and Georges Seurat. This colorimetric translation methodology is described and an example shown for the Chicago version of Vincent van Gogh's Bedroom. The methodology is compared with pixel‐based processing.     

采用color pilot软件处理数码照片评价牙齿的颜色变化

目的采用color pilot和Matlab软件处理数码照片及定点计算,评价数码照片是否能察觉牙齿颜色的变化。方法采用Vita比色板A2及C1作为试件,登士柏树脂附送的A2比色板作为参照,应用同一台数码照相机拍摄数码照片。根据光源、拍摄者、光圈、拍摄时间分组。将每张数码照片采用colorpilot软件处理,进行色彩矫正。结果经过计算机校正处理,同一试件的色差值基本上在1.5以下,有少数几项在1.5～3之间。其中,使用同一光圈的色差基本上都在1.5以下,不同的试验者,不同的时间,不同的照明环境对色差的影响并不明显。不同的光圈数值对色差的影响较大。不同的试件不论在何种条件下色差值均在3以上。结论经过色彩校正的数码照片可以察觉不同物体的颜色变化。     

Irradiance‐independent camera color calibration

For a digital color camera to represent the colors in the environment accurately, it is necessary to calibrate the camera RGB outputs in terms of a colorimetric space such as the CIEXYZ or sRGB. Assuming that the camera response is a linear function of scene luminance, the main step in the calibration is to determine a transformation matrix M mapping data from linear camera RGB to XYZ. Determining M is usually done by photographing a calibrated target, often a color checker, and then performing a least‐squares regression on the difference between the camera's RGB digital counts from each color checker patch and their corresponding true XYZ values. To measure accurately the XYZ coordinates for each patch, either a completely uniform lighting field is required, which can be hard to accomplish, or a measurement of the illuminant irradiance at each patch is needed. In this article, two computational methods are presented for camera color calibration that require only that the relative spectral power distribution of the illumination be constant across the color checker, while its irradiance may vary, and yet resolve for a color correction matrix that remains unaffected by any irradiance variation that may be present. © 2013 Wiley Periodicals, Inc. Col Res Appl, 39, 540–548, 2014     

Methods for designing characterisation targets for digital cameras

Characterisation targets usually include a set of physical coloured samples. A characterisation model can be derived between the colorimetric values (tristimulus values) and camera responses (RGB values) taken from an imaging device such as a digital camera capturing the colours in the target. The performance of such a model is highly dependent upon the number of colours and the colour region in the characterisation target. An ideal characterisation target should provide accurate model prediction without requiring too many samples. In this paper, a computational method is presented for colour selections to train a camera characterisation model based on a fourth‐order polynomial model including 35 terms. Compared with other available methods, the newly developed method performed better. It is proposed that this method be applied to generate generic targets in terms of colorimetric values. These targets should work reasonably well for a wide range of materials.     

Artificial neural networks for optimising camera‐based colour measurements of prints enhanced with pearlescent pigments

Pearlescent pigments are widely used in printing due to their optical, chemical and physical properties. To analyse the effects of goniochromism they produce, the colorimetric characterisation of materials printed with pearlescent pigments requires multi‐angular measurements. In this study, the colours of prints enhanced with pearlescent pigments were measured by means of a digital camera, relying on the empirical camera characterisation method. Since this method is time‐consuming, it was altered to enable estimates of colorimetric values for different geometries to be measured on the basis of images captured at one viewing angle. This approach was based on the use of artificial neural networks which were shown to provide sufficient flexibility for the given task. The results indicate that the images obtained at the viewing angle of 45° aspecular (measuring geometry 45°/asp 45°) accurately estimate CIELab values for all of the tested measuring geometries. The proposed method is therefore not only time‐efficient but also reduces the associated errors due to the camera's movement, and enables the estimation of colorimetric values for those viewing angles inaccessible by camera.     

Two‐primary crosstalk model for characterizing liquid crystal displays

The conventional methods for colorimetric characterization of displays assume that the displays satisfy the constraints of primary chromaticity invariance across gray levels and primary channel independence. The liquid crystal displays (LCDs) that reasonably satisfy the two constraints have been accurately characterized with the conventional methods and black‐level correction. For the LCDs that do not reasonably satisfy the two constraints, we propose a higher‐order method for accurate colorimetric characterization. Two‐primary crosstalk (TPC) is observed for two tested LCDs that may be due to signal interference. We derive the crosstalk function and develop the TPC model for characterizing the LCDs, which comprises a set of the simultaneous equations with offset constants, one‐color variables, and two‐color‐product variables. The results show that the accuracy of the TPC model is significantly improved compared with conventional device models and only slightly worse than the three‐dimensional look‐up‐table (3D‐LUT) model, while the numbers of measurement data are 49 and 512 for the TPC and 3D‐LUT models, respectively. The average color difference of 224 test samples is about 2.0 (1976 CIELAB color difference formula) with the TPC model for the LCD monitor either with higher or with lower two‐primary crosstalk. While the proposed TPC model yields improved characterization accuracy over conventional models, the TPC model is evaluated on only two LCDs of the same manufacturer. Thus, the generality of the LCD crosstalk deficiency is unknown and should be determined in future research. © 2006 Wiley Periodicals, Inc. Col Res Appl, 31, 90–101, 2006; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/col.20187     

Web‐based interactive dynamics for color models learning

A World Wide Web (WWW) based tool for the learning of color models is presented, along with a methodology of use. The system was developed in Java, allowing its remote use and a high degree of interaction with the user. It shows two color systems (RGB and HSB), users choose their position relative to the solids that define the system and, positioning control points, define the section of the solid to be designed. Both the system and the methodology are desktop‐oriented, and are low‐to‐moderately demanding in terms of both hardware and software. © 2000 John Wiley & Sons, Inc. Col Res Appl, 25, 435–441, 2000     

Calibrated color mapping between LCD and CRT displays: A case study

The primary goal of a color characterization model is to establish a mapping from digital input values d_i (i = R,G,B) to tristimulus values such as XYZ. A good characterization model should be fast, use a small amount of data, and allow for backward mapping from tristimulus to d_i. The characterization models considered here are for the case of an end user who has no direct knowledge of the internal properties of the display device or its device driver. Three characterization models tested on seven different display devices are presented. The characterization models implemented in this study are a 3D look up table (LUT) (Raja Balasubramanian, Reducing the Cost of Lookup Table Based Color Transformations, Proc IS&T/SID 7th Color Imaging Conference 1 ), a linear model (Fairchild MD, Wyble DR. Colorimetric Characterization of the Apple Studio Display (Flat Panel LCD). Munsell Color Science Laboratory Technical Report, 1998), and the masking model (Tamura N, Tsumura N, Miyake. Masking Model for accurate colorimetric characterization of LCD. Proc IS&T/SID 10th Color Imaging Conference 3 ). The devices include two CRT monitors, three LCD monitors, and two LCD projectors. The results of this study indicate that a simple linear model is the most effective and efficient for all devices used in the study. A simple extension to the linear model is presented, and it is demonstrated that this extension improves white prediction without causing significant errors for other colors. © 2005 Wiley Periodicals, Inc. Col Res Appl, 30, 438–447, 2005; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/col.     

Derivation of a color space for image color difference measurement

In digital image reproduction, it is often desirable to compute image difference of reproductions and the original images. The traditional CIE color difference formula, designed for simple color patches in controlled viewing conditions, is not adequate for computing image difference for spatially complex image stimuli. Zhang and Wandell [Proceedings of the SID Symposium, 1996; p 731–734] introduced the S‐CIELAB model to account for complex color stimuli using spatial filtering as a preprocessing stage. Building on S‐CIELAB, iCAM was designed to serve as both a color appearance model and also an image difference metric for complex color stimuli [IS&T/SID 10th Color Imaging Conference, 2002; p 33–38]. These image difference models follow a similar image processing path to approximate the behavior of human observers. Generally, image pairs are first converted into device‐independent coordinates such as CIE XYZ tristimulus values or approximate human cone responses (LMS), and then further transformed into opponent‐color channels approximating white‐black, red‐green, and yellow‐blue color perceptions. Once in the opponent space, the images are filtered with approximations of human contrast sensitivity functions (CSFs) to remove information that is invisible to the human visual system. The images are then transformed back to a color difference space such as CIELAB, and pixel‐by‐pixel color differences are calculated. The shape and effectiveness of the CSF spatial filters used in this type of modeling is highly dependent on the choice of opponent color space. For image difference calculations, the ideal opponent color space would be both linear and orthogonal such that the linear filtering is correct and any spatial processing on one channel does not affect the others. This article presents a review of historical opponent color spaces and an experimental derivation of a new color space and corresponding spatial filters specifically designed for image color difference calculations. © 2010 Wiley Periodicals, Inc. Col Res Appl, 2010     

Digital image‐color conversion between different illuminants by color‐constancy actuation in a color‐vision model based on the OSA‐UCS system

In digital image capture, the camera signals produced by the D65 illuminant, once translated into tristimulus values of the CIE 1931 standard colorimetric observer (assuming the Maxwell‐Ives‐Luther criterion is satisfied), are considered good to produce accurate color rendering. An image obtained under any illuminant other than D65 does not appear realistic and the tristimulus values of the camera must be transformed into the corresponding ones produced by the D65 illuminant. This transformation must satisfy color constancy. In this work, the transformation is obtained by a color‐vision model based on the Optical Society of America‐Uniform Color Scales system [Color Res Appl 2005; 30: 31–41] and is represented by a matrix dependent on the adaptation illuminant. This matrix is obtained by minimizing the distance between the pairs of the uniform scale chromatic responses related to the tristimulus values of the 99 different color samples of the SG Gretag‐Macbeth ColorChecker measured under a pair of different illuminants, one of which is the D65. Then any picture captured under a given light source can be translated into the picture of the same scene under the D65 illuminant. Metameric reason allows only approximate solutions. The transformations from Daylight and Planckian illuminants to the D65 illuminant have a very regular dependence on the color temperature, that appears to be the typical parameter for the color conversion. © 2012 Wiley Periodicals, Inc. Col Res Appl, 38, 412–422, 2013     

感光元件CMOS的灵敏度特性分析

数码相机的感光元件是相机的核心部位,是成像质量的关键影响因素。利用相机在不同条件下拍摄的照片,通过专业的图像处理软件分析照片的数据特点,并利用这些数据建立曝光量和灰度值之间的关系,对比分析不同相机、不同感光度等方面对感光特性曲线的影响。探讨了感光器件CMOS在不同条件下的感光规律。结果表明,利用数码相机拍照时应遵循多曝光不如少曝光的原则。     

Effects of cold plasma on the color parameters of Hyssop (Hyssopus officinalis L.) using color imaging instrumentation and spectrophotometer

This research was conducted to evaluate the effects of cold atmospheric plasma treatment on the color of Hyssop (Hyssopus officinalis L.) and also to compare the usage of the spectrophotometer vs the color imaging instrumentation for the evaluation of the treatment on the color parameters. The experiments were investigated at different treatment times of 1, 5, and 10 minutes and the voltage values of 17, 20, and 23 kV. Possible changes of color were evaluated by using CIE L*a*b* values obtained with HunterLab colorimeter and CIE L*a*b* values obtained with a digital still camera (DSC) using digital image processing (MATLAB software). The values of L*, a*, and b* of the samples were obtained using both the methods. The results revealed that the L*, a*, and b* values of the treated Hyssop samples changed with increasing the treatment time and the voltage applied. Evaluating the interaction effects revealed that there was a significant difference in the (−a^*/b^* ) ratio. In addition, the results showed that the effects of all variables on the color parameters were significantly different in the case of the DSC using digital image processing. However, these effects were not significantly different using HunterLab colorimeter except for time variable and interaction effects of a* and (−a^*/b^* ) ratio. The lightest green color and the maximum chlorophyll content loss were observed for 23 kV applied over 10 minutes. Based on the results, the digital image processing can be used as a practical tool to study the variations at the color of dried Hyssop leaves after cold plasma treatment.     

Color prediction models for digital Jacquard woven fabrics

The current industry practice for producing jacquard fabrics uses computer‐aided design (CAD) systems that provide visual simulations of the final color appearance of actual fabrics prior to production. This digital process is fundamentally based on the prediction of combined weave‐color effects, which can be successfully achieved by accurate color mixing models and the structural details of the fabrics. With the accurate models used in CAD systems, designers would see simulations more closely resembling fabrics to be produced. By checking the previews, the designers can easily modify, that is, recolor, the designs on the display monitor without doing repetitive physical sampling with the adjustment of the weaves and the yarn colors. However, there is no ready applicable accurate color mixing model for woven structures and there has not been sufficient investigation of the color prediction despite its usefulness for the current digital CAD process. Our study investigated the, color prediction of jacquard woven fabrics designed based on the principle of optically subtractive color mixing with the use of CMY colors. The color prediction was firstly done through the application of the six color mixing models previously developed for various other applications including fiber blending and printing. The performance of each model was evaluated by calculating the difference between the predicted and the measured colorimetric data, using ΔE_CMC(2:1). The average color difference from the models was 11.93 ΔE_CMC(2:1), which is hardly acceptable in textile industry. In order to increase the accuracy in color prediction, the six models were then optimized. As a result, substantial improvements for all models were obtained with a decrease in color difference to 4.83 ΔE_CMC(2:1) on average after the optimizations. Among the six optimized color mixing models, the optimized Warburton‐Oliver model, that is, W‐O model, was found to have the lowest average ΔE_CMC(2:1) value of approximately equaling to 2, which is considered potentially useful to be applied to the current digital fabric color prediction. © 2015 Wiley Periodicals, Inc. Col Res Appl, 41, 64–71, 2016     

Estimating spectral reflectance from camera responses based on CIE XYZ tristimulus values under multi‐illuminants

A new spectral reflectance estimation method based on CIE XYZ values under multi‐illuminants was proposed to obtain multi‐spectral images accurately by using digital still cameras. CIE XYZ values under multi‐illuminants were initially predicted from raw RGB responses by using a polynomial model with local training samples. Then, spectral reflectance was constructed from the predicted CIE XYZ values via the pseudo‐inverse method. Experimental results indicated that the new spectral reflectance estimation method significantly outperformed the traditional colorimetric characterization method without requiring extra training samples or greatly increasing computational complexities. © 2016 Wiley Periodicals, Inc. Col Res Appl, 42, 68–77, 2017