Color Constancy Algorithms Research Articles

Skin cancer extraction with optimum fuzzy thresholding technique

This paper discusses a new approach to segment different types of skin cancers using fuzzy logic approach. The traditional skin cancer segmentation involves the analysis of image features to delineate the cancerous region from the normal skin. Using low level features such as colour and intensity, segmentation can be done by obtaining a threshold level to separate the two regions. Methods like Otsu optimisation provide a quick and simple process to optimise such threshold level; however this process is prone to the lighting and skin tone variations. Fuzzy clustering algorithm has also been widely used in image processing due to its ability to model the fuzziness of human visual perception. Classical fuzzy C means (FCM) clustering algorithm has been applied to image segmentation with good results; however, the classical FCM is based on type-1 fuzzy sets and is unable to handle uncertainties in the images. In this paper, we proposed an optimum threshold segmentation algorithm based on type-2 fuzzy sets algorithms to delineate the cancerous area from the skin images. By using the 3D colour constancy algorithm, the effect of colour changes and shadows due to skin tone variation in the image can be significantly reduced in the preprocessing stage. We applied the optimum thresholding technique to the preprocessed image over the RGB channels, and combined individual results to achieve the overall skin cancer segmentation. Compared to the Otsu algorithm, the proposed method is less affected by the shadows and skin tone variations. The results also showed more tolerance at the boundary of the cancerous area. Compared with the type-1 FCM algorithm, the proposed method significantly reduced the segmentation error at the normal skin regions.

Human object annotation for surveillance video forensics

A system that can automatically annotate surveillance video in a manner useful for locating a person with a given description of clothing is presented. Each human is annotated based on two appearance features: primary colors of clothes and the presence of text/logos on clothes. The annotation occurs after a robust foreground extraction stage employing a modified Gaussian mixture model-based approach. The proposed pipeline consists of a preprocessing stage where color appearance of an image is improved using a color constancy algorithm. In order to annotate color information for human clothes, we use the color histogram feature in HSV space and find local maxima to extract dominant colors for different parts of a segmented human object. To detect text/logos on clothes, we begin with the extraction of connected components of enhanced horizontal, vertical, and diagonal edges in the frames. These candidate regions are classified as text or nontext on the basis of their local energy-based shape histogram features. Further, to detect humans, a novel technique has been proposed that uses contourlet transform-based local binary pattern (CLBP) features. In the proposed method, we extract the uniform direction invariant LBP feature descriptor for contourlet transformed high-pass subimages from vertical and diagonal directional bands. In the final stage, extracted CLBP descriptors are classified by a trained support vector machine. Experimental results illustrate the superiority of our method on large-scale surveillance video data.

Effects of chromatic image statistics on illumination induced color differences

We measure the color fidelity of visual scenes that are rendered under different (simulated) illuminants and shown on a calibrated LCD display. Observers make triad illuminant comparisons involving the renderings from two chromatic test illuminants and one achromatic reference illuminant shown simultaneously. Four chromatic test illuminants are used: two along the daylight locus (yellow and blue), and two perpendicular to it (red and green). The observers select the rendering having the best color fidelity, thereby indirectly judging which of the two test illuminants induces the smallest color differences compared to the reference. Both multicolor test scenes and natural scenes are studied. The multicolor scenes are synthesized and represent ellipsoidal distributions in CIELAB chromaticity space having the same mean chromaticity but different chromatic orientations. We show that, for those distributions, color fidelity is best when the vector of the illuminant change (pointing from neutral to chromatic) is parallel to the major axis of the scene's chromatic distribution. For our selection of natural scenes, which generally have much broader chromatic distributions, we measure a higher color fidelity for the yellow and blue illuminants than for red and green. Scrambled versions of the natural images are also studied to exclude possible semantic effects. We quantitatively predict the average observer response (i.e., the illuminant probability) with four types of models, differing in the extent to which they incorporate information processing by the visual system. Results show different levels of performance for the models, and different levels for the multicolor scenes and the natural scenes. Overall, models based on the scene averaged color difference have the best performance. We discuss how color constancy algorithms may be improved by exploiting knowledge of the chromatic distribution of the visual scene.

The biological significance of color constancy: An agent-based model with bees foraging from flowers under varied illumination

The perceived color of an object depends on its spectral reflectance and the spectral composition of the illuminant. Thus when the illumination changes, the light reflected from the object also varies. This would result in a different color sensation if no color constancy mechanism is put in place-that is, the ability to form consistent representation of colors across various illuminants and background scenes. We explore the quantitative benefits of various color constancy algorithms in an agent-based model of foraging bees, where agents select flower color based on reward. Each simulation is based on 100 "meadows" with five randomly selected flower species with empirically determined spectral reflectance properties, and each flower species is associated with realistic distributions of nectar rewards. Simulated foraging bees memorize the colors of flowers that they have experienced as most rewarding, and their task is to discriminate against other flower colors with lower rewards, even in the face of changing illumination conditions. We compared the performance of von Kries, White Patch, and Gray World constancy models with (hypothetical) bees with perfect color constancy, and color-blind bees. A bee equipped with trichromatic color vision but no color constancy performed only ∼20% better than a color-blind bee (relative to a maximum improvement at 100% for perfect color constancy), whereas the most powerful recovery of reflectance in the face of changing illumination was generated by a combination of von Kries photoreceptor adaptation and a White Patch calibration (∼30% improvement relative to a bee without color constancy). However, none of the tested algorithms generated perfect color constancy.

An evolutionary approach for colour constancy based on gamut mapping constraint satisfaction

Many colour constancy algorithms have been proposed to achieve a good performance in this field. The gamut mapping algorithm is one of the most accurate and promising algorithms based on gamut assumption: illuminant can be estimated by comparing the colours distribution in the current image to acanonical gamut (i.e., a pre-learned distribution of colours). However, the gamut mapping algorithm is precise but it is time consuming. Therefore, some other methods such as GCIE (gamut constrained illuminant estimation) and CGM (cubical gamut mapping) have been proposed which work faster. However, the results of such methods are poor when the source light is not in the pre-defined ones. In this paper, we propose an evolutionary algorithm for colour constancy based on gamut mapping assumption. This approach overcomes the mentioned problem in other gamut-based methods. The proposed evolutionary method uses a simple chromosome structure together with simple operators such as mutation, selection, and reproduction. Two versions of the proposed methods have been presented here. The first one works on image pixels, while the second one tries on image derivative. The experiments were done on three different data sets that are used in literature and results were satisfactory. The results showed that the proposed method is much improved when compared to other related methods in most of the time especially in the case of real world images.

Fast and robust color constancy algorithm based on grey block-differencing hypothesis

Color constancy is a fundamental research topic in color and vision. Numerous methods have been proposed in recent years. New methods are highly accurate but tend to be more complex. This paper proposes a simple low-level statistical algorithm based on a new hypothesis, the grey block-differencing hypothesis, which states that the average of reflectance differences of adjacent blocks in a scene is achromatic. The new method has almost the same complexity as the simplest methods (i.e., grey world and max-RGB). Experimental results demonstrate that the accuracy of the proposed method is exceptional.

Target recognitions in multiple-camera closed-circuit television using color constancy

People tracking in crowded scenes from closed-circuit television (CCTV) footage has been a popular and challenging task in computer vision. Due to the limited spatial resolution in the CCTV footage, the color of people’s dress may offer an alternative feature for their recognition and tracking. However, there are many factors, such as variable illumination conditions, viewing angles, and camera calibration, that may induce illusive modification of intrinsic color signatures of the target. Our objective is to recognize and track targets in multiple camera views using color as the detection feature, and to understand if a color constancy (CC) approach may help to reduce these color illusions due to illumination and camera artifacts and thereby improve target recognition performance. We have tested a number of CC algorithms using various color descriptors to assess the efficiency of target recognition from a real multicamera Imagery Library for Intelligent Detection Systems (i-LIDS) data set. Various classifiers have been used for target detection, and the figure of merit to assess the efficiency of target recognition is achieved through the area under the receiver operating characteristics (AUROC). We have proposed two modifications of luminance-based CC algorithms: one with a color transfer mechanism and the other using a pixel-wise sigmoid function for an adaptive dynamic range compression, a method termed enhanced luminance reflectance CC (ELRCC). We found that both algorithms improve the efficiency of target recognitions substantially better than that of the raw data without CC treatment, and in some cases the ELRCC improves target tracking by over 100% within the AUROC assessment metric. The performance of the ELRCC has been assessed over 10 selected targets from three different camera views of the i-LIDS footage, and the averaged target recognition efficiency over all these targets is found to be improved by about 54% in AUROC after the data are processed by the proposed ELRCC algorithm. This amount of improvement represents a reduction of probability of false alarm by about a factor of 5 at the probability of detection of 0.5. Our study concerns mainly the detection of colored targets; and issues for the recognition of white or gray targets will be addressed in a forthcoming study.

Analysis of colour constancy algorithms using the knowledge of variation of correlated colour temperature of daylight with solar elevation

In this article, we present an investigation of possible improvement of the colour constant reflectance features that can be obtained from daylight illuminated scenes using pixel-level colour constancy algorithms (model-based algorithm: S Ratnasingam, S Collins, J. Opt. Soc. Am. A 27, 286–294 (2010) and Projection-based algorithm: GD Finlayson, MS Drew, IEEE ICCV, 2001, pp. 473–480). Based on the investigation we describe a method to improve the performance of the colour constancy algorithms using the correlation between the correlated colour temperature of measured daylight with the solar elevation and phase of the day (morning, midday and evening). From this observation, the data from 1 year are used to create a solar elevation and phase of day-dependant method of interpreting the information obtained the colour constancy algorithms. Test results show that using the proposed method with 40-dB signal-to-noise ratio the performance of the projection-based algorithm and model-based algorithm can be improved on average by 33.7 and 45.4%, respectively. More importantly, a larger improvement (85.9 and 113.7%) was obtained during the middle period of each day which is defined as when the solar elevation is larger than 20°.

Color constancy algorithm without segmentation

Color constancy algorithms are often used for illuminant compensation. This paper proposes an illuminant compensation method using a camera noise analysis without segmentation. Pixels within highlight regions generally include large amounts of information on the image illuminant. Thus, the analysis of highlight regions provides a relatively easy means of determining the characteristics of an image illuminant. Given that sensor non-uniformity affects illuminant estimation, the proposed method selects valid pixels within a highlight region by analyzing the chromatic noise characteristics of the camera. The illuminant is then estimated using a modified dichromatic reflectance model based on the selected valid pixels. Experimental results confirm that the proposed method is superior to other conventional methods.

Balanced color contrast enhancement for digital images

The balanced enhancement for color images refers to the ability of improving the visual quality of both highlights and dark areas. To this end, a symmetric Naka-Rushton formula (SNRF) is proposed. Particularly, SNRF is symmetric around a pixel-wise threshold which is devised as the adaptive criterion for highlights. Treating RGB channels separately, SNRF regulates each pixel according to the anisotropic local intensity, thus avoiding halos and protecting uniform areas. Further, the color-constancy extension of SNRF (SNRFCC) is presented by incorporating conventional image statistics with the adaptive threshold. Different from traditional color-constancy algorithms that focus on color correction, SNRFCC accomplishes color and lightness adaptation simultaneously. Finally, a postprocessing technique is designed to compensate for the slightly compressed contrast due to the balanced adaptation. Compared with other tone-mapping methods, SNRF achieves more vivid color and details especially in highlights. Besides, comparisons with other color-contrast enhancement algorithms confirm that SNRFCC features the merit of less residual color cast when implementing color constancy on images under varying light.

Color constancy by chromaticity neutralization

In this paper, a robust illuminant estimation algorithm for color constancy is proposed. Considering the drawback of the well-known max-RGB algorithm, which regards only pixels with the maximum image intensities, we explore the representative pixels from an image for illuminant estimation: The representative pixels are determined via the intensity bounds corresponding to a certain percentage value in the normalized accumulative histograms. To achieve the suitable percentage, an iterative algorithm is presented by simultaneously neutralizing the chromaticity distribution and preventing overcorrection. The experimental results on the benchmark databases provided by Simon Fraser University and Microsoft Research Cambridge, as well as several web images, demonstrate the effectiveness of our approach.

REMOVING SHADOWS FROM HIGH-RESOLUTION URBAN AERIAL IMAGES BASED ON COLOR CONSTANCY

Abstract. A method is explored to remove tall building shadows in true color and color infrared urban aerial images based on the theory of color constancy. This paper first uses the specthem ratio and Otsu threshold segmentation methods to detect building shadows on urban aerial true color and color infrared aerial images. Then, based on the shadow detection result, one of the color constancy algorithms SoG (Shades of Gray) is used to remove the shadows in aerial images with different p values of the Minkowski norm. Finally, the shadow removal results with different p values have been compared by brightness, contrast and average gradients. The experiments show that the result of this method based on color constancy has a good visual effect, and different from general scene image shadow removal, the aerial images get the best shadow removal result when p is 2. It means the two types of aerial images should not be simply regarded as gray world images.

Experiments on the White Patch Retinex in RGB and CIELAB color spaces

Color constancy is an important process in a number of vision tasks. Most devices for capturing images operate on the RGB color space and, usually, the processing of the images is in this space, although some processes have shown a better performance when a perceptual color space is used instead. In this paper, experiments on the White Patch Retinex, a color constancy algorithm commonly used, are performed in two color spaces, RGB and CIELAB, for comparison purposes. Experimental results using an imagery set are analyzed using a no-reference quality metric and outcomes are discussed. It has been found that the White Patch Retinex algorithm shows a better performance in RGB than in CIELAB, but when color adjustments are implemented in sequence, firstly in CIELAB and then in RGB, much better results are obtained.

The Role of Bright Pixels in Illumination Estimation

The White-Patch method, one of the very first colour constancy methods, estimates the illuminant colour from the maximum response of three colour channels. However, this simple method has been superseded by advanced physical, statistical and learning based colour constancy methods. Recently, a few research works have suggested that the simple idea of using maximum pixel values is not as limited an idea as it seems on first glance. These works show that in several situations some manipulations can indeed made it perform very well. Here, we extend the White-Patch assumption to include any of: white patch, highlights or light source; let us refer to these pixels in an image as the “bright” pixels areas. We propose that bright pixels are surprisingly helpful in the illumination estimation process.In this paper, we investigate the effects of bright pixels on several current colour constancy algorithms. Moreover, we describe a simple framework for an illumination estimation method based on bright pixels and compare its accuracy to well-known colour constancy algorithms applied to four standard datasets. We also investigate failure and success cases, using bright pixels, and propose desiderata on input images with regard to the proposed method.

Adaptive colour constancy algorithm using discrete wavelet transform

The colours of chromatically homogeneous object surfaces measured by a sensor vary with the illuminant colour used to illuminate the objects. In contrast, colour constancy enables humans to identify the true colours of the surfaces under varying illumination. This paper proposes an adaptive colour constancy algorithm which estimates the illuminant colour from wavelet coefficients at each scale of the decomposition by discrete wavelet transform of the input image. The angular error between the estimated illuminant colours in consecutive scales are used to determine the optimum scale for the best estimate of the true illuminant colour. The estimated illuminant colour is then used to modify the approximation subbands of the image so as to generate the illuminant-colour corrected image via inverse discrete wavelet transform. The experiments show that the colour constancy results generated by the proposed algorithm are comparable or better than those of the state-of-the-art colour constancy algorithms that use low-level image features.

Color based skin classification

Skin detection is used in applications ranging from face detection, tracking body parts and hand gesture analysis, to retrieval and blocking objectionable content. In this paper, we investigate and evaluate (1) the effect of color space transformation on skin detection performance and finding the appropriate color space for skin detection, (2) the role of the illuminance component of a color space, (3) the appropriate pixel based skin color modeling technique and finally, (4) the effect of color constancy algorithms on color based skin classification. The comprehensive color space and skin color modeling evaluation will help in the selection of the best combinations for skin detection. Nine skin modeling approaches (AdaBoost, Bayesian network, J48, Multilayer Perceptron, Naive Bayesian, Random Forest, RBF network, SVM and the histogram approach of Jones and Rehg (2002)) in six color spaces (IHLS, HSI, RGB, normalized RGB, YCbCr and CIELAB) with the presence or absence of the illuminance component are compared and evaluated. Moreover, the impact of five color constancy algorithms on skin detection is reported. Results on a database of 8991 images with manually annotated pixel-level ground truth show that (1) the cylindrical color spaces outperform other color spaces, (2) the absence of the illuminance component decreases performance, (3) the selection of an appropriate skin color modeling approach is important and that the tree based classifiers (Random forest, J48) are well suited to pixel based skin detection. As a best combination, the Random Forest combined with the cylindrical color spaces, while keeping the illuminance component outperforms other combinations, and (4) the usage of color constancy algorithms can improve skin detection performance.

Color Constancy for Multiple Light Sources

Color constancy algorithms are generally based on the simplifying assumption that the spectral distribution of a light source is uniform across scenes. However, in reality, this assumption is often violated due to the presence of multiple light sources. In this paper, we will address more realistic scenarios where the uniform light-source assumption is too restrictive. First, a methodology is proposed to extend existing algorithms by applying color constancy locally to image patches, rather than globally to the entire image. After local (patch-based) illuminant estimation, these estimates are combined into more robust estimations, and a local correction is applied based on a modified diagonal model. Quantitative and qualitative experiments on spectral and real images show that the proposed methodology reduces the influence of two light sources simultaneously present in one scene. If the chromatic difference between these two illuminants is more than 1°, the proposed framework outperforms algorithms based on the uniform light-source assumption (with error-reduction up to approximately 30%). Otherwise, when the chromatic difference is less than 1° and the scene can be considered to contain one (approximately) uniform light source, the performance of the proposed method framework is similar to global color constancy methods.

Color constancy by characterization of illumination chromaticity

Computational color constancy algorithms play a key role in achieving desired color reproduction in digital cameras. Failure to estimate illumination chromaticity correctly will result in invalid overall colour cast in the image that will be easily detected by human observers. A new algorithm is presented for computational color constancy. Low computational complexity and low memory requirement make the algorithm suitable for resource-limited camera devices, such as consumer digital cameras and camera phones. Operation of the algorithm relies on characterization of the range of possible illumination chromaticities in terms of camera sensor response. The fact that only illumination chromaticity is characterized instead of the full color gamut, for example, increases robustness against variations in sensor characteristics and against failure of diagonal model of illumination change. Multiple databases are used in order to demonstrate the good performance of the algorithm in comparison to the state-of-the-art color constancy algorithms.

Color Constancy Using Natural Image Statistics and Scene Semantics

Existing color constancy methods are all based on specific assumptions such as the spatial and spectral characteristics of images. As a consequence, no algorithm can be considered as universal. However, with the large variety of available methods, the question is how to select the method that performs best for a specific image. To achieve selection and combining of color constancy algorithms, in this paper natural image statistics are used to identify the most important characteristics of color images. Then, based on these image characteristics, the proper color constancy algorithm (or best combination of algorithms) is selected for a specific image. To capture the image characteristics, the Weibull parameterization (e.g., grain size and contrast) is used. It is shown that the Weibull parameterization is related to the image attributes to which the used color constancy methods are sensitive. An MoG-classifier is used to learn the correlation and weighting between the Weibull-parameters and the image attributes (number of edges, amount of texture, and SNR). The output of the classifier is the selection of the best performing color constancy method for a certain image. Experimental results show a large improvement over state-of-the-art single algorithms. On a data set consisting of more than 11,000 images, an increase in color constancy performance up to 20 percent (median angular error) can be obtained compared to the best-performing single algorithm. Further, it is shown that for certain scene categories, one specific color constancy algorithm can be used instead of the classifier considering several algorithms.

Extending “color constancy” outside the visible region

In this paper, the results of an investigation of the possibility of extending "color constancy" to obtain illuminant-invariant reflectance features from data in the near-ultraviolet (UV) and near-infrared (IR) wavelength regions are reported. These features are obtained by extending a blackbody-model-based color constancy algorithm proposed by Ratnasingam and Collins [J. Opt. Soc. Am. A27, 286 (2010)] to these additional wavelengths. Ratnasingam and Collins applied the model-based algorithm in the visible region to extract two illuminant-invariant features related to the wavelength-dependent reflectance of a surface from the responses of four sensors. In this paper, this model-based algorithm is extended to extract two illuminant-invariant reflectance features from the responses of sensors that cover the visible and either the near-UV or near-IR wavelength. In this investigation, test reflectance data sets are generated using the goodness-fitness coefficient (GFC). The appropriateness of the GFC for generating the test data sets is demonstrated by comparing the results obtained with these data with those obtained from data sets generated using the CIELab distance. Results based upon the GFC are then presented that suggest that the model-based algorithm can extract useful features from data from the visible and near-IR wavelengths. Finally, results are presented that show that, although the spectrum of daylight in the near UV is very different from a blackbody spectrum, the algorithm can be modified to extract useful features from visible and near-UV wavelengths.