On Finding Gray Pixels

Abstract

We propose a novel grayness index for finding gray pixels and demonstrate its effectiveness and efficiency in illumination estimation. The grayness index, GI in short, is derived using the Dichromatic Reflection Model and is learning-free. GI allows to estimate one or multiple illumination sources in color-biased images. On standard singleillumination and multiple-illumination estimation benchmarks, GI outperforms state-of-the-art statistical methods and many recent deep methods. GI is simple and fast, written in a few dozen lines of code, processing a 1080p image in ∼ 0.4 seconds with a non-optimized Matlab code.

Fast Fourier Color Constancy and Grayness Index for ISPA Illumination Estimation Challenge

“…This is shown by its mean error, as several "outlier" images can bring very high angular errors, increasing the mean error rapidly. This is consistent with the observations in [8]. Figure 1 illustrates several hard images in the testing set for GI.…”

“…Grayness Index [8] addresses color constancy from a different angle. Apply log and laplacian-of-gaussian filter δ on Equation.…”

“…We submitted the results given by this model P. GI-based method We did not change anything -we downloaded the code from the github page 1 and used it on testing images. Hard samples from the undisclosed testing set, which can fail statistical methods like GI [8], while learning-based methods (FFCC [3] and the ranked-1st CNN method) can perform much better. Images are gammacorrected (gamma 2.2) for better visual effects.…”

“…The other submission is ranked 6th which is out of top three, but we still want to introduce it. It is based on Grayness Index (GI) [8], which is an advanced version of gray pixel methods [9], [14]. A comparison of gray pixel methods is given by [10].…”

Flash Lightens Gray Pixe

“…Our method relies on gray pixel detection [13]. The original work assumes Lambertian surfaces, and then revisited and improved by [16,17]. The original and extended gray pixel methods however fail in the case of mixed illumination (the top row in Fig.…”