Search results for "Color constancy"
showing 3 items of 13 documents
Improving color correction across camera and illumination changes by contextual sample selection
2012
International audience; In many tasks of machine vision applications, it is important that recorded colors remain constant, in the real world scene, even under changes of the illuminants and the cameras. Contrary to the human vision system, a machine vision system exhibits inadequate adaptability to the variation of lighting conditions. Automatic white bal- ance control available in commercial cameras is not sufficient to pro- vide reproducible color classification. We address this problem of color constancy on a large image database acquired with varying digi- tal cameras and lighting conditions. A device-independent color repre- sentation may be obtained by applying a chromatic adaptation…
<title>Selective filters for improvement of color discrimination</title>
2003
Colored filters have been used as possible aids for perception in case of congenital color vision abnormalities. Higher stimulation for L cones in one eye (red filter) and for M cones (green) in other eye acts as a simple method to interfere in human color vision processing. Influence on subjective and objective vision performances as binocular functions, fusion, stereopsis and color vision has been tested. Special colored glasses have been created, taking into account each subjects refraction for distance and tested for the 5 months adaption time. No negative influence on binocular functions, no changes of heterophoria and lowering in fusion ability and stereopsis have been stated due to u…
ROBUST ROAD SIGNS SEGMENTATION IN COLOR IMAGES
2012
International audience; This paper presents an efficient method for road signs segmentation in color images. Color segmentation of road signs is a difficult task due to variations in the image acquisition conditions. Therefore, a color constancy algorithm is usually applied prior to segmentation, which increases the computation time. The proposed method is based on a log-chromaticity color space which shows good invariance properties to changing illumination. Thus, the method is simple and fast since it does not require color constancy algorithms. Experiments with a large dataset and comparison with other approaches, show the robustness and accuracy of the method in detecting road signs in …