Fast underwater color correction using integral images

Underwater image processing has to face the problem of loss of color and contrast that occurs when images are acquired at a certain depth and range. The longer wavelengths of sunlight such as red or orange are rapidly absorbed by the water body, while the shorter ones have a higher scattering. There...

Descripción completa

Detalles Bibliográficos
Autores: Neumann, László, Garcia, Rafael, Jánosik, József, Gracias, Nuno
Tipo de recurso: artículo
Fecha de publicación:2018
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/126282
Acceso en línea:https://hdl.handle.net/2117/126282
Access Level:acceso abierto
Palabra clave:Underwater imaging systems
Underwater imaging
Image enhancement
Color correction
Comunicacions subacuàtiques
Fons marins -- Investigació
Àrees temàtiques de la UPC::Enginyeria electrònica
Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal::Processament de la imatge i del senyal vídeo
Descripción
Sumario:Underwater image processing has to face the problem of loss of color and contrast that occurs when images are acquired at a certain depth and range. The longer wavelengths of sunlight such as red or orange are rapidly absorbed by the water body, while the shorter ones have a higher scattering. Thereby, at larger distance, the scene colors appear bluish-greenish, as well as blurry. The loss of color increases not only vertically through the water column, but also horizontally, so that the subjects further away from the camera appear colorless and indistinguishable, suffering from lack of visible details. This paper presents a fast enhancement method for color correction of underwater images. The method is based on the gray-world assumption applied in the Ruderman-opponent color space and is able to cope with non-uniformly illuminated scenes. Integral images are exploited by the proposed method to perform fast color correction, taking into account locally changing luminance and chrominance. Due to the low-complexity cost this method is suitable for real-time applications ensuring realistic colors of the objects, more visible details and enhanced visual quality.