Nonlinear Color Space Research Articles

Dynamic Selective Edge-Based Integer/Fractional-Order Partial Differential Equation for Degraded Document Image Binarization

Conventional thresholding algorithms have had limited success with degraded document images. Recently, partial differential equations (PDEs) have been applied with good results. However, these are usually tailored to handle relatively few specific distortions. In this study, we combine an edge detection term with a linear binarization source term in a PDE formulation. Additionally, a new proposed diffusivity function further amplifies desired edges. It also suppresses undesired edges that comprise bleed-through effects. Furthermore, we develop the fractional variant of the proposed scheme, which further improves results and provides more flexibility. Moreover, nonlinear color spaces are utilized to improve binarization results for images with color distortion. The proposed scheme removes document image degradation such as bleed-through, stains, smudges, etc., and also restores faded text in the images. Experimental subjective and objective results show consistently superior performance of the proposed approach compared to the state-of-the-art PDE-based models.

Nonlinear color space coded by additive digital pulses

Additive color mixing modulated by digital pulses enables universal applications in lights and displays. Conventional methods map binarized signals to ordinary red–green–blue color spaces, loosely connected with the color perception of human eyes, causing the complexity of gamut mapping and inaccuracy of chromatic manipulations. Here we developed a complete theory that encodes and decodes digital signals directly to a perceptually nonuniform color space of the commission internationale de l’éclairage, featuring a strict bijection between the duty cycles of each binary pulse and color components of the mixed light and an analytic nonlinear gamut volume. Exemplary applications in biophotonic lighting, active full-color displaying, and metaphotonic sensing confirmed our theory. The method applies to the fields with accurate manipulations of chromaticity and luminosity of a light, opening an avenue toward the next generation of perceptual displays, cameras, and sensors.

FPGA-Based Multiplier-Less Log-Based Hardware Architectures for Hybrid Color Image Enhancement System

Logarithmic image processing (LIP) models provide a viable framework for image enhancement, particularly the color-based models. These techniques, due to their vector nature, avoid the problems of grayscale methods applied to RGB color image processing. Furthermore, the absence of multiply-accumulate (MAC) operations makes them amenable to hardware realization for faster operation when compared with spatial filter kernel-based approaches. However, details of image scenes are not always revealed after processing and homomorphic filters are still a popular low-complexity option for correction of images acquired under uneven illumination conditions, notwithstanding the fading that occurs when processing RGB color images. Though nonlinear color space converter hardware architectures have been implemented and reported in the literature for improving results from conventional illumination correction techniques, they add to the hardware complexity. The contributions of the work described in this paper includes original, practical, standard and logarithmic, multiplier-less high-speed architectures for a variant of the LIP model in addition to standard and log-hybrid architectures for a modified spatial homomorphic filter (MSDHF) to reduce color fading in RGB space. The modified homomorphic filter is able to enhance both faded/bright and dark low-contrast images, giving it an edge over standard illumination correction systems. A combined approach utilizing the MSDHF and simplified LIP algorithms enhances the strengths of the individual algorithms while minimizing their individual weaknesses. Comparisons of the developed systems are made with the standard homomorphic filter architecture and similar works from the literature.

A Second-Order TV-Type Approach for Inpainting and Denoising Higher Dimensional Combined Cyclic and Vector Space Data

In this paper we consider denoising and inpainting problems for higher dimensional combined cyclic and linear space valued data. These kind of data appear when dealing with nonlinear color spaces such as HSV, and they can be obtained by changing the space domain of, e.g., an optical flow field to polar coordinates. For such nonlinear data spaces, we develop algorithms for the solution of the corresponding second order total variation (TV) type problems for denoising, inpainting as well as the combination of both. We provide a convergence analysis and we apply the algorithms to concrete problems.

Robust Periocular Recognition by Fusing Sparse Representations of Color and Geometry Information

In this paper, we propose a re-weighted elastic net (REN) model for biometric recognition. The new model is applied to data separated into geometric and color spatial components. The geometric information is extracted using a fast cartoon - texture decomposition model based on a dual formulation of the total variation norm allowing us to carry information about the overall geometry of images. Color components are defined using linear and nonlinear color spaces, namely the red-green-blue (RGB), chromaticity-brightness (CB) and hue-saturation-value (HSV). Next, according to a Bayesian fusion-scheme, sparse representations for classification purposes are obtained. The scheme is numerically solved using a gradient projection (GP) algorithm. In the empirical validation of the proposed model, we have chosen the periocular region, which is an emerging trait known for its robustness against low quality data. Our results were obtained in the publicly available UBIRIS.v2 data set and show consistent improvements in recognition effectiveness when compared to related state-of-the-art techniques.

Variability estimation of hue and saturation components in the HSV space

When dealing with hue-saturation-value (HSV) components, one must be very careful about nonlinearity and singularities of the HSV space, which can amplify considerably the sampling errors of the original components, namely, red-green-blue (RGB) channels. Contradictorily, almost all research in computerized image analysis or processing based on HSV (or other nonlinear color space) neglect to account for such inhomogeneous variability. In this article, we present two approaches for estimating the degree of variability of hue and saturation in the HSV space, intended to assist any color image processing or analysis based on such components. Both approaches end up with new estimators of the hue and saturation standard deviation for a pixel distribution of any color sample. The first approach consists in applying error-propagation analysis on the RGB-to-HSV formulation and the second one is based on heuristic rules inferred from geometrical relations between HSV values of pixel distributions. Experiments show that introduced estimators provide very accurate predictions of true H-S deviations obtained from a broad subset of 183 NCS® color samples. © 2011 Wiley Periodicals, Inc. Col Res Appl, 2011

Detection of Skin Erythema in Darkly Pigmented Skin Using Multispectral Images

To develop a technique using a fixed, discrete set of wavelengths that can detect erythema in persons with darkly pigmented skin. The resulting erythema detection approach will then be incorporated into a handheld, point-of-care device that is clinically viable and affordable. A multispectral imaging system was used to acquire spectral images of induced erythema. Individual images were combined into a single image using different fusion algorithms. Image fusion algorithms based on published literature and using linear and nonlinear color space transformation were tested to optimize the contrast between erythematic and uninvolved skin. A research laboratory at Georgia Institute of Technology, Atlanta, Georgia. Fifty-six subjects, of whom 28 had darkly pigmented skin, were recruited from a pool of students, faculty, and staff. The ability of detection algorithms to detect erythema was measured using Weber contrast. A simple threshold classifier determined accuracy, sensitivity, and specificity for each algorithm. Four algorithms enhanced contrast of erythema by an order of magnitude over that of a digital photograph. The accuracy of the detection algorithms ranged from 66% to 95%. Sensitivity and specificity ranged from 0% to 100%. One fusion algorithm exhibited an accuracy of more than 90% and sensitivity and specificity of more than 90%. The results indicate that erythema in different skin tones can be identified using 2 to 3 filters. Increasing accuracy and discrimination will be targeted via use of filters with narrower half-wave bandwidths, more consistent camera lighting, and improved machine vision techniques.

Designing effective transfer functions for volume rendering from photographic volumes

Photographic volumes present a unique, interesting challenge for volume rendering. In photographic volumes, the voxel color is pre-determined, making color selection through transfer functions unnecessary. However, photographic data does not contain a clear mapping from the multi-valued color values to a scalar density or opacity, making projection and compositing much more difficult than with traditional volumes. Moreover, because of the nonlinear nature of color spaces, there is no meaningful norm for the multi-valued voxels. Thus, the individual color channels of photographic data must be treated as incomparable data tuples rather than as vector values. Traditional differential geometric tools, such as intensity gradients, density and Laplacians, are distorted by the nonlinear non-orthonormal color spaces that are the domain of the voxel values. We have developed different techniques for managing these issues while directly rendering volumes from photographic data. We present and justify the normalization of color values by mapping RGB values to the CIE L*u*v* color space. We explore and compare different opacity transfer functions that map three-channel color values to opacity. We apply these many-to-one mappings to the original RGB values as well as to the voxels after conversion to L*u*v* space. Direct rendering using transfer functions allows us to explore photographic volumes without having to commit to an a-priori segmentation that might mask fine variations of interest. We empirically compare the combined effects of each of the two color spaces with our opacity transfer functions using source data from the Visible Human project.

カラーマネジメントおよび色彩工学 構造解析的手法に基づく均等色空間構成方法

Color-information representation by using a uniform color space in terms of human JNDs (just-noticeable differences) is important in many industrial situations as well as in image processing, visually loss-less image coding, and noise evaluation. In this paper, we describe a new method for building a non-linear color space, which makes use of various JND data sets. We evaluated our algorithm by computer simulation and found that the color space obtained by the algorithm was superior to that obtained by conventional methods in terms of uniformity and representation compactness.