Translator Disclaimer
18 August 2011 Visibility enhancement for robust tracking under bad weather
Author Affiliations +
In this paper, we present a novel method for visibility enhancement based on atmospheric scattering imaging models. Given only a single degraded image, we firstly estimate global atmospheric light vector based on dark channel prior. Then fast bilateral filter is used to deduce atmospheric veil, which is the key contribution of this paper. Following these, the ideal scene radiance could be recovered by directly solving physics-based imaging equation finally. The main advantage of our weather removal algorithm is that, it does not require any a priori scene structure, distributions of scene reflectance, or detailed knowledge about the particular weather condition, and could achieve similar or better restoration results with only a fraction of time consumption in contrast to state-of-art techniques both for color and grey images. Experiments results demonstrate that out algorithm could significantly enhance the details of hazy images, which is very important for features extraction and robust tracking for out-door vision system.
© (2011) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Sun Kang and Wang Bo "Visibility enhancement for robust tracking under bad weather", Proc. SPIE 8194, International Symposium on Photoelectronic Detection and Imaging 2011: Advances in Imaging Detectors and Applications, 819438 (18 August 2011);


Back to Top