A method to reproduce images on an object under various observation geometries is presented. In this method, a multispectral image sequence is captured with rotating the object under a point light source, where spectral distribution and the position are measured. The diffuse and specular reflection images are decomposed from the captured images based on the dichromatic reflection and the Lambertian models. The angular distribution of reflected light is obtained by the decomposed images, and the images under new observation geometry are synthesized by using light-ray rearrangement technique. In the experiments using two types of 2D objects, leather and fabric, it is confirmed that synthesized images under new illumination geometry are almost the same as the images actually captured under the new geometry.

Electronic printing is now 25 years old. A great deal of progress has been made as evidenced by the great number of electronic printers and their associated technologies. This paper illustrates three simple metrics with which to analyze electronic printer technology. The metrics are not necessarily intended to be anything profound or deterministic but appear to illustrate some relative constants as well as significant improvements in printers that indicate a trend whose final outcome is yet to be seen. Since electronic printers are basically information delivery devices, they can be analyzed as to their information delivery capability on a cost-performance basis. The three metrics utilized in this paper involve a number of characteristics such as printer weight, speed, spatial printing density and color. Over 170 printers area used in this analysis dating from the birth of the laser printer in the mid-1970s up to the present time. This metric oriented view of the technology may give us some insight as to where the electronic printing industry is going on a cost/performance basis as well as where new products should be aimed so that they provide the desired market impact.

Traditional technologies for image capture devices are based on the three RGB filters for the generation of the color signals. As an alternative, in the last years spectral methods of image capture were introduced with the prospect of several advantages for image quality and color data handling. Most of these solutions are so-called multi- spectral technologies and therefore still use filters. In this paper a spectral approach for image capturing is discussed, which is based on a diffraction grating. The grating is integrated on a micro-mirror device in a linear array structure with very small dimensions. The main design criteria and the important parameters are discussed. Also the influences of MEMS technologies for the production of such devices are shown.

The quality control of fabrics makes still use of visual assessments in cases where the highest degree of matching between batch and reference samples must be achieved, e.g. in the field of 'mix and match' garments. The reasons for the failure of current color measuring instruments were investigated, and it was found that among others, the most important aspects were the vivid nature of a textile whose color cannot be completely described by just a single measurement, and the differences of the conditions of viewing and measuring. A multispectral scanner which was originally designed for faithfully digitizing originals such as textiles and paintings was used to measure the colors of fabric samples. Since the spectral scanner was not limited to a standard measuring geometry, assimilation of the viewing and measuring geometries was possible. The system was run through a field test in the quality control with a large manufacturer of fabrics. It turned out that, unlike with conventional spectrophotometers, a very good congruence between the pass/fail decisions of the multispectral system and experienced persons could be achieved.

Traditional image processing techniques used for 3- and 4- band images are not suited to the many-band character of spectral images. A sparse multi-dimensional lookup table with inter-node interpolation is a typical image processing technique used for applying either a known model or an empirically derived mapping to an image. Such an approach for spectral images becomes problematic because input dimensionality of lookup tables is proportional to the number of source image bands and the size of lookup table sis exponentially related to the number of input dimensions. While an RGB or CMY source images would require a 3D lookup table, a 31-band spectral image would need a 31-dimensional lookup table. A 31-dimensional lookup table would be absurdly large. A novel approach to spectral image processing is explored. This approach combines a low-cost spectral analysis followed by application of one from a set of low-dimensional lookup tables. The method is computationally feasible and does not make excessive demands on disk space or run-time memory.

We report a technique for spectral image compression to be used in the field of data communications. The spectral domain of the images is represented by a low-dimensional component image set, which is used to obtain an efficient compression of the high-dimensional spectral data. The component images are compressed using a similar technique as the JPEG- and MPEG-type compressions use to subsample the chrominance channels. The spectral compression is based on Principal Component Analysis (PCA) combined with color image transmission coding technique of 'chromatic channel subsampling' of the component images. The component images are subsampled using 4:2:2, 4:2:0, and 4:1:1-based compressions. In addition, we extended the test for larger block sizes and larger number of component images than in the original JPEG- and MPEG-standards. Totally 50 natural spectral images were used as test material in our experiments. Several error measures of the compression are reported. The same compressions are done using Independent Component Analysis and the results are compared with PCA. These methods give a good compression ratio while keeping visual quality of color still good. Quantitative comparisons between the original and reconstructed spectral images are presented.

For dry toner electrophotographic color printers, we present a numerical simulation model describing the color printer responses based on a physical characterization of the different electrophotographic process steps. The proposed model introduces a Cross Transfer Efficiency designed to predict the color transmittance spectra of multi-color prints by taking into account the transfer influence of each deposited color toner layer upon the other layers. The simulation model leads to a better understanding of the factors that have an impact on printing quality. In order to avoid the additional optical non-linearities produced by light reflection on paper, we have limited the present investigation to transparency prints. The proposed model succeeded to predict the transmittance spectra of printed wedges combining two color toner layers with a mean deviation less than CIE-LAB (Delta) E equals 2.5.

Due to the increasing popularity and afford ability of color imaging devices, color characterization for these devices becomes an important subject. In other words, a set of color profile(s) needs to be generated for each device to transform the device dependent color space to a device independent one. This paper will concentrate on color characterization of scanners.

ICC profile specification defines data format of color device characteristics to realize the accurate color reproduction. However, device characteristics are not constant on various conditions such as temperature, humidity, aging and so on. It may cause deterioration of the device independency. We have developed new extended profile structure based on ICC profile to compensate color deviation by sing feedback signal of variable factors. The effectiveness of the prosed method has been confirmed through experiments of color ink-jet printer. Proposed method can be implemented as an extension of existing ICC profile.

A key problem in multimedia systems is the faithful reproduction of color. One of the main reasons why this is a complicated issue are the different color reproduction technologies used by the various devices; displays use easily modeled additive color mixing, while printers use a subtractive process, the characterization of which is much more complex than that of self-luminous displays. In order to resolve these problems several processing steps are necessary, one of which is accurate device characterization. Our study examines different learning algorithms for one particular neural network technique which already has been found to be useful in related contexts, namely radial basis function network models, and prosed a modified learning algorithm which improves the colorimetric characterization process of printers. In particular our results show that it is possible to obtain good performance by using a learning algorithm that is trained on only small sets of color samples, and use it to generate a larger lookup table through use of multiple polynomial regression or an interpolation algorithm. We deem our findings to be a good start point for further studies on learning algorithms, used in conjunction with this problem.

Printing with custom inks is useful for extending the gamut of printed images, for creating artistically appealing design or for providing protection against counterfeiting. The basic colors we consider, consist of the custom inks, their superpositions and the white paper. Color separation for custom inks requires to determine the relative amounts of the basic colors allowing to render each desired input color. To achieve this goal, one may tetrahedrize the set of basic colors on a given 3D color space. However, even for a few basic colors, there are many different tetrahedrizations.

Ink limit is an important parameter for printer color calibration, especially for inkjet printers. A GCR approach is often used to control the total ink amount for CMYK printers. However, a tradition GCR approach has the following limitations: 1) it can not reduce the total ink amount to less than 200 percent for CMYK printers; 2) it can not be applied to reduce ink for CMY printers; 3) to achieve highest image quality, ink amount may be limited to different values in different regions, in which the GCR approach fails. In this paper, a new approach is presented to control ink limit. It controls ink limit globally as well as locally. An algorithm was developed to construct a gamut boundary for gamut mapping that guarantees that the constructed gamut surface covers only colors within the ink limit. If the ink limit needs to be modified, the gamut surface is reconstructed based on the original measured data. Therefore redoing and remeasuring a target is avoided. It greatly simplifies the ink limit control and color calibration.

A device gamut can be visualized as a 3D object in a given color space. The globe is approximately a 3D ellipsoid and has been the motivation for centuries of research in the area of geometric projections. Neither the globe nor a typical device gamut is flat but there are often times when it is useful to examine the globe and device gamuts in 2D. Cartography provides a multitude of geometric projections and terminology that can be used for gamut visualization and mapping. This paper provides examples of how the concepts and principles of cartography can be used for gamut visualization and gamut mapping. The Mercator, Cassini, Bonne, Mollweide, and Sinusoidal projections are used to visualize the sRGB gamut. Finally, a simple example using spherical coordinates is used to demonstrate how gamut projection can be used to implement a specific type of gamut mapping.

A new target for defining media gamut boundaries is described. The use of measured gamut boundary data using provided by this target is capable of improving the accuracy of gamut boundary predictions in both device model and gamut mapping computations. The accuracy of existing methods of calculating gamut boundaries can be improved by use of an enlarged gamut boundary training set to derive the gamut boundary descriptor. An alternate, computationally simple, method of calculating gamut boundary intersections is described, and its performance in calculating such intersections and in media mapping is described. There is good agreement between this method and FSLGB in predicting boundary intersections if the same data is used to derive the gamut boundary descriptor. The gamut boundary target described in this paper is thus proposed as a means of obtaining greater information about media gamut boundaries.

The color gamuts of color imaging media are important parameters in the reproduction of color images between them and their assumed magnitudes directly influence the degree to which colors are modified. In spite of this, the determination of gamut boundaries is often done in a way that ignores some basic implications that follow from the definition of color gamuts. This is partly due to the fact that some of these implications are not understood and partly due to the fact that if they are understood their magnitude is underestimated. Hence, the approach that is taken in this paper is to first discuss the theoretical implications of what color gamuts are and subsequently to illustrate them by experimental means..

This paper describes a simple research and pedagogical tool for thinking about color gamut compromises. The idea is to fabricate Mondrian-like images that have patches of different colors with intermediate spatial complexity between single stimuli and real complex images. These gamut Mondrians can be used to illustrate the observations found in the literature and to act as a simple experimental system to explore the principles of gamut mapping algorithms.

Gamut mapping is a very actual problem in today's color reproduction. For a target device, a gamut mapping algorithm establishes a correspondence between the out of gamut color and the color within the gamut, trying to preserve same color appearance with the original color when that color is rendered on the target device. This problem becomes more important with the increasing number of color reproduction devices and cross media image representation. This paper focuses on some issues related to gamut mapping. First, the optimization of the target gamut is discussed because it may reduce many of the gamut mapping problems by reducing the volume of the out of gamut colors. Second, the influence of the color space in which the gamut mapping performs is discussed and a linearization procedure for CIELAB space is proposed. Third, a short overview of the existing gamut mapping procedure is given.

Gamut mapping algorithms were tested in a transparency-to- newsprint workflow, using features derived from studies of empirical mappings in high-quality color reproductions. The experiment compared different methods of determining the achromatic convergence point in simulations lightness-chroma compression, and compared linear compression against a non- linear distance-weighted compression. Algorithms whose convergence points were dependent on the lightness and chroma of the cusp and of the color being mapping performed better than those with fixed convergence points. The models using non-linear compression were strongly preferred over one using linear compression.

Digital cameras are no longer exotic gadgets being used by a privileged group of early adopters. More and more people realize that there are obvious advantages to the digital solution over the conventional film-based workflow. Claiming that prints on paper are no longer necessary in the digit workflow, however, would be similar to reviving the myth of the paperless office. Often, people still like to share their memories on paper and this for a variety of reasons. There are still some hurdles to be taken in order to make the digital dream com true. In this paper, we will give a survey of the different workflows in digital photography. The local, semi-local and Internet solutions will be discussed as well as the preferred output systems for each of these solutions. When discussing output system, we immediately think of appropriate color management solutions. In the second part of this paper, we will discuss the major color management issues appearing in digital photography. A clear separation between the image acquisition and the image rendering phases will be made. After a quick survey of the different image restoration and enhancement techniques, we will make some reflections on the ideal color exchange space; the enhanced image should be delivered in this exchange space and, from there, the standard color management transformations can be applied to transfer the image from this exchange space to the native color space of the output device. We will also discus some color gamut characteristics and color management problems of different types of photographic printers that can occur during this conversion process.

Digital Still Cameras employ automatic white balance techniques to adjust sensor amplifier gains so that white imaged objects appear white. A color cast detection algorithm is presented that uses histogram and segmentation techniques to select near-neutral objects in the image. Once identified and classified, these objects permit determination of the scene illuminant and implicitly the respective amplifier gains. Under certain circumstances, a scene may contain no near-neutral objects. By using the segmentation operations on non-neutral image objects, memory colors, from skin, sky, and foliage objects, may be identified. If identified, these memory colors provide enough chromatic information to predict the scene illuminant. By combining the approaches from near-neutral objects with those of memory color objects, a reasonable automatic white balance over a wide range of scenes is possible.

Chromatic adaptation transforms are used in imaging system to map image appearance to colorimetry under different illumination sources. In this paper, the performance of different chromatic adaptation transforms (CAT) is compared with the performance of transforms based on RGB primaries that have been investigated in relation to standard color spaces for digital still camera characterization and image interchange. The chromatic adaptation transforms studied are von Kries, Bradford, Sharp, and CMCCAT2000. The RGB primaries investigated are ROMM, ITU-R BT.709, and 'prime wavelength' RGB. The chromatic adaptation model used is a von Kries model that linearly scales post-adaptation cone response with illuminant dependent coefficients. The transforms were evaluated using 16 sets of corresponding color dat. The actual and predicted tristimulus values were converted to CIELAB, and three different error prediction metrics, (Delta) E_Lab, (Delta) E_CIE94, and (Delta) E_CMC(1:1) were applied to the results. One-tail Student-t tests for matched pairs were calculated to compare if the variations in errors are statistically significant. For the given corresponding color data sets, the traditional chromatic adaptation transforms, Sharp CAT and CMCCAT2000, performed best. However, some transforms based on RGB primaries also exhibit good chromatic adaptation behavior, leading to the conclusion that white-point independent RGB spaces for image encoding can be defined. This conclusion holds only if the linear von Kries model is considered adequate to predict chromatic adaptation behavior.

Most color appearance models that have been published so far require a simplified description of the viewing field which is subdivided into a small number of homogeneous regions. The tristimulus values and luminance levels of these regions serve as input parameters for the models. The purpose of this paper however is to study brightness perception in a complex, achromatic surround using a more detailed description of the entire viewing field. Therefore, a number of psycho physical experiments were carried out using a CRT display on which relatively complex images were presented. Several observers were asked to judge the perceived brightness by adjusting the luminance level of a reference grey for a perfect brightness match. All thereby obtained psycho physical data were used to develop a new brightness appearance model that takes all objects in the entire visual field into account. The model includes an object feature extraction stage, in which object properties like area and position are extracted, a stage in which the characteristic object data are sorted into the appropriate classes of a histogram and a multivariate map in the form of a feedforward neural network to calculate the prediction of the perceived brightness.

We have proposed colormaps to replace a widely used 216 colors palette in order to allow a designer with normal color vision to simulate the colors seen by dichromats. As dichromats lack one class of cone photo pigment, they confuse colors that differ only in the excitation of the missing class of photo pigment. The method is based on the LMS colorimetric system, which specifies colors in terms of the relative excitations of the cones. We have constructed a rule to reduce any set of confused colors to a single three- component specification. We have introduced a modification, assuming that the video display primaries and nominal white are representative of recent standards for Cathode Ray Tube monitors and that its video-transfer function is a power function with an exponent of 2.2. For everyday practice, replacing a normal palette by a reduced palette provides an immediate warning of possible losses of readability of a display by color-deficients.

Photorealistic Image Synthesis is a relevant research and application field in computer graphics, whose aim is to produce synthetic images that are undistinguishable from real ones. Photorealism is based upon accurate computational models of light material interaction, that allow us to compute the spectral intensity light field of a geometrically described scene. The fundamental methods are ray tracing and radiosity. While radiosity allows us to compute the diffuse component of the emitted and reflected light, applying ray tracing in a two pass solution we can also cope with non diffuse properties of the model surfaces. Both methods can be implemented to generate an accurate photometric distribution of light of the simulated environment. A still open problem is the visualization phase, whose purpose is to display the final result of the simulated mode on a monitor screen or on a printed paper. The tone reproduction problem consists of finding the best solution to compress the extended dynamic range of the computed light field into the limited range of the displayable colors. Recently some scholars have addressed this problem considering the perception stage of image formation, so including a model of the human visual system in the visualization process. In this paper we present a working hypothesis to solve the tone reproduction problem of synthetic image generation, integrating Retinex perception model into the photo realistic image synthesis context.

The reproduction of natural objects in color images has attracted a great deal of attention. Reproduction more pleasing colors of natural objects is one of the methods available to improve image quality. We developed an automatic color correction method to maintain preferred color reproduction for three significant categories: facial skin color, green grass and blue sky. In this method, a representative color in an object area to be corrected is automatically extracted from an input image, and a set of color correction parameters is selected depending on the representative color. The improvement in image quality for reproductions of natural image was more than 93 percent in subjective experiments. These results show the usefulness of our automatic color correction method for the reproduction of preferred colors.

Various rendering techniques are being used for document reproduction and printing. Some rendering techniques work better for text, some for graphics and some others work better for picture regions. Therefore, dividing a document image into regions that need to be rendered differently from its neighboring regions is useful for good reproduction and printing. In this paper we describe a method to classify previously segmented regions of a page image into three classes, namely text, graphics and pictures. In addition to printing and copying, this classification of regions into broad basic classes is also useful for automatic storage and retrieval, and efficient communication of document images.

Anew color image segmentation algorithm is presented in this paper. This algorithm is invariant to highlights and shading. This is accomplished in two steps. First, the average pixel intensity is removed form each RGB coordinate. This transformation mitigates the effects of highlights. Next, the Mixture of Principal Components algorithm is used to perform the segmentation. The MPC is implicitly invariant to shading due to the inner vector product or vector angle being used as similarity measure. Since the new coordinate system contains negative numbers, it is necessary to modify the MPC algorithm since in its original form it does not distinguish between positive and negative color space coordinates. Results on artificial and real images illustrate the effectiveness of the method. Finally, the use of the total within-cluster variance is investigated as possible criterion for selecting the number of clusters for the new algorithm.

In an electronic color imaging device such as a digital camera using a single CCD or CMOS sensor, the color information is usually acquired in sub-sampled patterns of red (R), green (G) and blue (B) pixels. Full resolution color is subsequently generated from this sub-sampled image. This is popularly called Color Interpolation or Color Demosaicing. In this paper, we present a color interpolation algorithm using the method of fuzzy membership assignment along with the concept of smooth hue transition. The algorithm is adaptive in nature and produces superior quality full resolution color images compared to most of the popularly known color interpolation algorithms in the literature. Performance of the algorithm has been compared with a previously proposed block matching algorithm for color interpolation by the authors as well as the popularly used bilinear color interpolation. We present the results of comparison with some challenging sub-sampled images for color interpolation.

We have developed a novel 1200 dot per inch light emitting diode (LED) printhead for a high quality digital color printer in the next generation. Performance tests showed that the LED printhead was really suitable for a color printer printhead: an excellent light output uniformity within +/- 1 percent, a high contrast light spot of 80 percent, a high accuracy light spot position within +/- 20 micrometers , and a very high light output efficiency to realize a high-speed printing of 45 pages per minute. In order to achieve the high performance of the LED printhead, we have developed an original 1200 dpi LED array chip with 8 by 8 matrix-structured wiring. The chip-matrix LED array chip was fabricated on a double hetero structured Al_xGa_1-xAs epitaxial wafer using our original fabrication techniques and device designs. This is the key for fabricating a super-high resolution array chip with a super- high light output efficiency. Moreover, the CM LED array chip made it possible to reduce an overall super-high- density printhead cost: the number of bonding wires and driver integrated circuit chips on the LED printhead decreases to one-fourth, and one-half, respectively, compared to the conventional 1200 dpi printhead.

The database AMPAC proposes the simple and unified format to describe single parameter of whole field of design, production and management. The database described by the format can be used commonly in any field connected by the network production system, since the description accepts any parameter in any fields and is field independent definition.

The feature analysis of the conventional exchange format used in the field of the graphic arts is performed. It becomes clear from this consideration that most standard defining the transmission format have too strict limitations to adapt for the communication required in the creative and flexible work.

In this paper, we describe the main feature of a system supporting the selection of color palettes for qualitative data representation and graphic interface design. The system is mainly based on visual interaction providing effective tools for browsing the Munsell color space and setting perceptual constraints on the number and type of colors the system selects automatically. The system can also manage ICC device profiles, making it possible to process colors in terms of standard, device-independent color representation. Experimental results are also reported.

The authors have developed a new color conversion system that can contribute to the improvement of imperfect color reproduction in various types of displays, especially those of non-CRT displays such as TFT-LCD monitors. This paper describes the new color conversion system and touches on modeling of the TFT-LCD monitors. The new system is a system that uses a new matrix calculation that we have developed. Owing to its ease in hardware application, this color conversion system allows support of motion picture display and real time processing. In addition, it enables realization of desired color conversion characteristics and it also facilitates realization of use to two and more types of color conversion characteristics by selection. The authors have also developed color conversion simulation software based upon high-accurate modeling of the TFT-LCD monitors which can be applied to a practical TFT-LCD monitor to set the parameters of new color conversion system to an optimal condition.

Techniques to evaluate image quality are necessary. However, several problems to be investigated still remain in the evaluation of the image quality. Though it is difficult to evaluate the image quality totally, the sharpness and graininess are focused on as components of total image quality. As examples of the evaluation and improvement of image quality, two experimental results are shown in this paper. At fist, the outline and remarks in the evaluation of image quality are described. Then experimental results about evaluation of multi-level error diffusion images and improvement of noisy images without sharpness degradation are presented. Finally future problems and the conclusions are addressed.

Color Management (CM) tools claim to give both accurate and consistent results. However using CM tools of distinct manufacturers, different results are likely to occur even if the same parameters are given. Obviously the manufacturers do not focus on the same criteria doing their optimizations, which leads to the fact that each CM tool has its own strengths and weaknesses. In this study, new methods for comparing CM tools were evaluated, and with these methods, ICC profiles generated by five CM tools of current leaders on the market were investigated in order to point out their weaknesses and strengths. In addition, the influence of using various CM modules (CMM) was tested. For the generation of each ICC profile, the same measurements were used. Since the standardized ISO 12642 file format, in which the measurement data are stored, was not accepted by each of the tools, the file had to be converted to proprietary formats. The investigated quality aspects were accuracy by each of the tools, the file had to be converted to proprietary criteria. For each aspect, performance was visualized by mapping the (Delta) E values to pseudo colors, giving a very intuitive view on the investigated subject.

We have already proposed a standard portrait for the assessment of preferable skin tone. The present report describes a psycho physical experimental method, i.e., simultaneous triplet comparison that has been developed for the assessment of skin tone by using the portrait and that is characterized not only by a scalability, stability and reproducibility of the resulting scale values, but also by a reduce stress on observers. We have confirmed that the present simultaneous triplet comparison has a degree of scalability and stability almost equivalent to that of paired comparison that is most widely used for similar purposes, and that the stress on observers is about half as much as that of paired comparison.

A new graininess metrics, differing from the traditional ones in taking into consideration the effects of lightness, chroma and hue angle, and applicable to digital prints, is posed. Print samples from electrophotography, offset print, ink jet, thermal dye transfer, thermofusible ink transfer, and silver halide are evaluated with this new metrics, and the results are compared. The graininess index calculated from this metrics is used to classify the graininess rating based on subjective observations.

The colors in printed products should be known before printing. There are now many inexpensive inkjet printers able to print color. Using color management systems they can be used as proofing devices. The color gamut and the stability of these devices were assessed. There are tow main results of this project: We have developed a procedure and the tools to qualify proofing devices and we have shown that there are inexpensive devices suited for proofing newspaper print. The gamut is wide enough to simulate offset printing on uncoated paper and it is possible to simulate the color tint of the paper on which the final product will be printed. It is possible to attain a small mean Delta E between such an inexpensive proof and the final printed product. This Delta E is often smaller than between an analog proof and the final printed product. The prerequisites are calibration of the printer and color profiles of both the printer and the printing process to be simulated.

A method for the quality management of paint producers was developed that allows for an objective description of inhomogeneous fading of paint coatings after free weathering using relevant metric quantities such as color contrast, gradient of color contrast, and geometric features of the inhomogeneous structures. These may be quantified with the method of digital color image analysis. The first step to apply this technique means a systematic investigation of the color transformation properties specific of the selected input/output devices used for digital imaging. To build a color management system mathematical models of the color transformation processes were optimized and embedded in a commercial color image analysis software. The needed metric parameters, that evaluate the damages on the coated surfaces, must be deduced for highest possible agreement with visual judgements of experts on the categorization of the damages. 150 samples of paint coatings after weathering were selected to investigate this correlation.

color images are usually viewed on CRT monitor screen under ambient light sources. The feeling for reproduced color on CRT monitor depends on white point of ambient light source and the CRT monitor screen. Therefore, if the effect of ambient light source is not considered, then the reproduced color does not match the intrinsic color. This paper proposes a method to investigate the effect of ambient light source on color appearance of a CRT monitor which is posited in a room illuminated by an experimental light source. The investigation for the effect of ambient light consists of two elements. One is the flare effect that the reflection on the CRT monitor screen which produced by the ambient light source. Another is the change of adaption in white point. When the images are viewed on a CRT monitor screen under ambient light, the sate of adaption is affected by both the white point of the monitor and ambient light source. Therefore, these two effects were analyzed in the experiments.

The direct binary search (DBS) algorithm is one of a family of iterative approaches to signal design. DBS is characterized by a particular search strategy and an efficient approach to evaluating the effect of trial changes. From its origins as a method for synthesis of digital holograms, we trace the evolution of DBS to its current state as a powerful and flexible method for digital half toning and a tool for development of derivative algorithms for digital half toning. We describe the theory that underpins how DBS behaves, and discuss in some detail the use of printer models within DBS. Finally, we describe briefly the application of DBS to color half toning and its use to design halftone screens. A lookup table based algorithm, aperiodic micro screens, and halftones with an embedded watermark.

Most commercially printed images are halftoned using a screening process. IN order to reproduce printed documents containing images, one usually performs descreening or inverse half toning to avoid possible moire patterns. There exists a variety of gray scale halftone descreening techniques in the literature. However, color halftone descreening is still an ongoing research topic. In this paper, we present two descreening approaches: suboptimal FIR filter and a two-stage color sigma filter. The suboptimal FIR descreening filter offers an efficient descreening approach for gray scale halftoned images. In the mean time, a color halftone descreening technique based on the color sigma filter does not assume any a priori knowledge about the half toning process, making it applicable to any color halftone image. Similar to the anisotropic diffusion algorithm and total variation minimization techniques designed for gray scale images, the color sigma filter is an O(N) algorithm which can smooth out variation within each region and preserve edge information in the RGB color space. When combined with halftone segmentation techniques, a complete document processing algorithm for gray-scale and color documents can be created.

Most of the half toning algorithms are based on ideal imaging devices that can render perfect square pixels. In real printing environments this is not the case. Most imaging deices are a trade-off between the best quality and the highest speed. In this paper a screen will be designed for Agfa's newspaper-dedicated computer-to-plate imaging device Polaris.

This paper introduces a moire-free color half toning algorithm. Moire patterns might appear in traditional halftone images. This could be caused by interference among halftone screens of different colors, or beating between he input image and the screens. In our algorithm, we adjust the halftone screens of different colors, or beating between the input image and the screens. In our algorithm, we adjust the halftone output in such a way that the low-frequency difference between the input and the appearance of the output is minimized. Due to non-linear color mixing, the appearance of a color print is not a simple combination of CMYK four channels, and is hence not easily predictable. We use the 2 by 2 printer model to estimate the pixel-level CIE-Lab values from the halftone output. Next we calculate the difference between the original CIE-Lab input image and the 2 by 2 estimation and run a low-pass filtering on the result. Then, we feed back the low-frequency difference into the input. The low-frequency-feedback process cancels the potential color moire patterns.

We address the periodic clustered tool dot color screen design problem. In traditional clustered dot color screening, the screen for each colorant is rotated to a different angle relative to the others. If the angles are not carefully chosen, visible moire and rosette artifacts may appear. These artifacts primarily result from the interaction of the periodic structures associated with the halftone screens of different colorants. Registration errors can also introduce unwanted artifacts in the screened images. Using lattice theory and a model for the perceived rendered halftone, we present a systematic method for designing moire and rosette free clustered dot color screens for discrete-raster color systems. We also investigate strategies for making the resulting screen robust to registration errors.

In this work, we take a microstructure model based approach to the problem of color prediction of halftones created using an inkjet printer. We assume absorption and scattering of light through the colorant layers and model the subsurface light scattering in the substrate by a Gaussian point spread function. We restrict our analysis to transparent substrates. To model the absorption and scattering of light through the colorant layers, we employ the Kubelka-Munk color mixing mode. To model the scattering in the substrate and to predict the spectral distribution, we use a wavelength dependent version of the reflection prediction model developed by Ruckdeschel and Hauser. Using spectral distributions and ink weight measurements for transparencies completely and homogeneously coated with colorants, we compute the absorption and scattering spectra of the colorants using the Kubelka-Munk theory. We train our model using measured spectral distribution and synthesized microstructure images of primary ramps printed on transparent media. For each patch in the primary ramp, we synthesize a high-resolution halftone microstructure image from the halftone bitmap assuming dot profiles with Gaussian roll-offs, form which we compute a high-resolution transmission image using the Kubelka-Munk theory and the absorption and scattering spectra of the colorants. We then convolve this transmission image with the Gaussian point spread function of the transparent substrate to predict the average spectral distribution of the halftone. We use our model to predict the spectral distribution of a secondary ramp printed on the same media.

A model for the human visual system (HVS) is an important component of many half toning algorithms. Using the iterative direct binary search (DBS) algorithm, we compare the halftone texture quality provided by four different HVS models that have been reported in the literature. Choosing one HVS model as the best for DBS, we then develop an approximation to that model which significantly improves computational performance while minimally increasing the complexity of the code. By varying the parameters of this mode, we find that it is possible to tune it to the gray level being rendered, and to thus yield superior halftone quality across the tone scale. We then develop a dual-metric DBS algorithm that effectively provides a tone-dependent HVS model without a large increase in computational complexity.

In this contribution, we introduce a multiple depth progressive representation for network-based still and moving images. A simple quantization algorithm associated with this representation provides optimal image quality. By optimum, we mean the best possible visual quality for a given value of information under real life constraints such as physical, psychological , or legal constraints. A special variant of the algorithm, multi-depth coherent error diffusion, addresses a specific problem of temporal coherence between frames in moving images. The output produced with our algorithm is visually pleasant because its Fourier spectrum is close to the 'blue noise'.

Several improved algorithms to map color information to color or black texture are introduced. The new algorithms are based on pattern mixing and halftoning. The detailed algorithms are described along with the comparative analysis of their implementation. The analysis and the experiment result show that the halftone based algorithms described in this paper have advantages over the pattern mixing algorithm.

Subjective image quality measurement usually resorts to psycho physical scaling. However, it is difficult to evaluate the inherent precision of these scaling methods. Without knowing the potential errors of the measurement, subsequent use of the data can be misleading. In this paper, the errors on scaled values derived form paired comparison based scaling methods are simulated with randomly introduced proportion of choice errors that follow the binomial distribution. Simulation results are given for various combinations of the number of stimuli and the sampling size. The errors are presented in the form of average standard deviation of the scaled values and can be fitted reasonably well with an empirical equation that can be sued for scaling error estimation and measurement design. The simulation proves paired comparison based scaling methods can have large errors on the derived scaled values when the sampling size and the number of stimuli are small. Examples are also given to show the potential errors on actually scaled values of color image prints as measured by the method of paired comparison.

A possibility of creation of a high-velocity laser printing machine of the 'computer to print' type is considered. The ink is removed from a transparent ink carrier on the image carrier under a pressure pulse arising of the UV laser radiation focusing point. The experimental results are reported on the removal of different-color ink form a transparent substrate on an image carrier by means of the VU laser pulse. It is shown that the energy required for printing a single element may amount to several hundred nanojoules.