Exemplar-Based Color Constancy and Multiple Illumination

Abstract

Exemplar-based learning or, equally, nearest neighbor methods have recently gained interest from researchers in a variety of computer science domains because of the prevalence of large amounts of accessible data and storage capacity. In computer vision, these types of technique have been successful in several problems such as scene recognition, shape matching, image parsing, character recognition, and object detection. Applying the concept of exemplar-based learning to the problem of color constancy seems odd at first glance since, in the first place, similar nearest neighbor images are not usually affected by precisely similar illuminants and, in the second place, gathering a dataset consisting of all possible real-world images, including indoor and outdoor scenes and for all possible illuminant colors and intensities, is indeed impossible. In this paper, we instead focus on surfaces in the image and address the color constancy problem by unsupervised learning of an appropriate model for each training surface in training images. We find nearest neighbor models for each surface in a test image and estimate its illumination based on comparing the statistics of pixels belonging to nearest neighbor surfaces and the target surface. The final illumination estimation results from combining these estimated illuminants over surfaces to generate a unique estimate. We show that it performs very well, for standard datasets, compared to current color constancy algorithms, including when learning based on one image dataset is applied to tests from a different dataset. The proposed method has the advantage of overcoming multi-illuminant situations, which is not possible for most current methods since they assume the color of the illuminant is constant all over the image. We show a technique to overcome the multiple illuminant situation using the proposed method and test our technique on images with two distinct sources of illumination using a multiple-illuminant color constancy dataset. The concept proposed here is a completely new approach to the color constancy problem and provides a simple learning-based framework.

Multi-Cue Illumination Estimation via a Tree-Structured Group Joint Sparse Representation

“…We compare our MC method with three new illumination estimation methods: double-opponency mechanisms-based method (DO) proposed by Gao et al (2013), exemplarbased method (EB) proposed by Joze and Drew (2014), and corrected-moment-based method (CMM) proposed by Finlayson (2013). Since the codes of these methods are not publicly available, we cannot evaluate their performance using our experimental settings described in Sect.…”

Deep structured-output regression learning for computational color constancy

“…Bayesian approaches [22] learn the probability of illumination assuming normal-distributed reflectances. The Exemplar method [29] estimates illumination via finding nearest neighbor surfaces of a test image using an unsupervised clustering of texture and color features,…”

Illuminant Chromaticity Estimation via Optimization of RGB Channel Standard Deviation

“…Grey world, Max-RGB or white patch, shades of grey [3] , first and second order grey edge [4] and bright and dark colors PCA [21] were some of the statistical-based color constancy methods used for the comparison. The representative learning-based color constancy methods included edge and pixel based gamut mapping [22] , bayesian based [23] , regression (SVR) [24] and exemplar based method [25] .…”