The authors present a comprehensive method of spectral reflectance estimation for points sampled on a 3D surface in a structured-light projection system. The proposed solution does not require any additional hardware apart from the devices used in the classic structured-light projection setup. It utilizes a calibrated color CCD camera and a DLP projector, which projects sine fringes in order to provide precise 3D point coordinates. Additionally, the projector is used as a light source for multispectral image capture by displaying uniform color fields. Two approaches are verified for the design of optimal projected hues. Simultaneous acquisition of the 3D shape information as well as the multispectral data is possible in a single sequence of images. They are corrected for illumination nonuniformity, ambient light and camera noise. The presented method reconstructs a point cloud representing the surface with spectral reflectance estimated for each point. It uses the Wiener inverse method for spectrum reconstruction. The authors provide an evaluation of the accuracy of the proposed method and show the results of scanning painting samples created with different techniques, as an example application of cultural heritage digitization.
Spectral information divergence (SID) was identified as the most efficient spectral similarity measure. However, we show that divergence are not adapted to direct use on spectra. Following an idea proposed by Nidamanuri, we construct a spectral pseudo-divergence based on the Kullback–Leibler divergence. This pseudo-divergence is composed of two parts: a shape and an intensity similarity measure. Consequently, bidimensional representation of spectral differences are constructed to display the histograms of similarity between a spectral reference and the spectra from a dataset or an hyperspectral image. We prove the efficiency of the spectral similarity measure and of bidimensional histogram of spectral differences on artificial and Cultural-Heritage spectral images.
Multichannel printing employs additional colorants to achieve higher quality reproduction, assuming their physical overlap restrictions are met. These restrictions are commonly overcome in the printing workflow by controlling the colorant choice at each point. Our multilevel halftoning algorithm bundles inks of same hues in one channel with no overlap, separating them into eight channels, consequentially benefitting of increased ink options at each point. In this article, implementation and analysis of the algorithm is carried out. Color separation is performed using the cellular Yule–Nielsen modified spectral Neugebauer model. The channels are binarized with the multilevel halftoning algorithm. The workflow is evaluated with an eight-channel inkjet at 600 dpi resulting in mean and maximum ΔE94 color differences around 1 and 2, respectively. The halftoning algorithm is analyzed using S-CIELAB, thus involving the human visual system, in which multilevel halftoning showed improvement in terms of image quality compared to the conventional approach.
The features of personal skin conditions are often perceived by how they appear on local skin areas. The appearance of a local skin surface depends on two features: the color texture caused by pigment distributions in skin layers and the brightness textures of sulcus cutis and shading. The image analysis of skin conditions requires the decomposition of local skin images into these components. This article proposes a novel method to estimate these local skin properties by referring to the multispectral images of a skin surface. Spectral images are decomposed into color and brightness textures through a multi-resolution analysis with wavelet functions. The pigment distributions are then estimated based on the Kubelka–Munk theory. Photorealistic images of skin with different local skin conditions can be generated by following the reverse process of decomposition based on the adjusted local texture components. The experimental results demonstrated the good feasibility of the proposed method.
In this article, the authors present a method for assessing image quality in stereoscopic images: QUALITAS. The proposed method is inspired by some features of the human visual system, such as contrast sensitivity, response to visual disparity and perception of distance. Individual qualities of the stereo-pair are not simply averaged. QUALITAS introduces Contrast Band-Pass Filtering on a wavelet domain in both views; in this way it weights left and right images perceptually depending on viewing conditions. The authors have tested the method on the LIVE 3D stereoscopic image database and compared the results with a wide set of image quality metrics from current research.
Most of digital cameras today use a color filter array (CFA) and a single sensor to acquire color information of the scene. In this article, we ask which arrangement of colors in the mosaic of the CFA provides the best encoding of the scene. As a solution of the inverse problem of demosaicing, we consider a linear minimum mean squared error model. We used redundancy given by the neighborhood on the sampled image to ensure the stability of the solution. For some CFAs, LMMSE with neighborhood provides equivalent reconstruction results and less variability among the image content compared to edge-directed demosaicing on the Bayer. LMMSE allows comparing CFAs of regular pattern with random ones. We show that mosaics with random arrangement of colors and quasi equal proportion of RGB provide best reconstruction performance.
In this article, we propose a joint halftoning and data hiding technique for color images. To ensure high quality of the printed image, the color direct binary search (CDBS) iterative halftoning algorithm is used. The proposed approach uses the commonly available cyan, magenta and yellow colorants to hide data in the chrominance channels. Orientation modulation is used for data embedding during the iterative CDBS halftoning stage. The detector is using PCA-learned components to extract the embedded data from the scanned image. Experimental results show that this proposed CDBS-based data hiding method offers both higher data hiding capacity and higher robustness to the print-and-scan channel when compared to the state-of-the-art grayscale counterpart method. The relatively high correct detection rate make this approach suitable for applications which require exact extraction of embedded data in prints.
Warmth perception is a physical, emotional, semantic, and sensorial bond between people and their environments. Although the effects of single colors have been explored, there has been no research on how paired colors affect warmth perception in interiors. Therefore, the main aim of this study is to investigate these effects of colors and color pairs. Each model was assessed by 32 participants, totaling 96 different participants assessed the color models (Red, White, Green, and their pairs) under controlled conditions, both on a seven-point semantic differential scale and through open-ended questions. The results show that both single colors and paired colors affect warmth perception in interiors. White, Green, and Red are warmer than each other, respectively. Red appears to increase and White appears to decrease the warmth perception of their pairs in interiors. Another important finding of the study is that there is no effect of color location in paired colors.
Color vision deficiency (CVD) simulation methods are used both to simulate color vision of color-deficient people and as input for image enhancement methods for the color-deficient. However, a standardized method to compare simulation methods has not yet been defined. We propose a behavioral methodology to evaluate, compare and rank different simulation methods. By using accuracy and response time data from a match-to-sample experiment, we can assess behavioral performance of simulation methods. We show firstly that the match-to-sample paradigm is well suited to show performance differences between observer groups; secondly, that the simulation methods do indeed simulate the desired CVD to some degree; and thirdly, we show that the proposed methodology can be used to rank different simulation methods. Our results indicate that the simulation method proposed by Brettel et al. depicts deutan CVD more accurately than the simulation method proposed by Kotera.