European Commission logo
español español
CORDIS - Resultados de investigaciones de la UE
CORDIS

Dataset and dehazing methods for non-homogeneous and dense hazy scenes

Descripción del proyecto

Un método mejorado para eliminar la niebla de imágenes

En condiciones climáticas desfavorables, como la niebla, la calidad de las imágenes se degrada bastante debido a la presencia de partículas en suspensión en el aire que dispersan la luz. Esto dificulta la identificación de objetos, bien por la visión humana, bien por sistemas de visión artificial. Las técnicas de eliminación de niebla existentes no son muy eficaces cuando se trata de imágenes con niebla densa y homogénea. El proyecto financiado con fondos europeos NH-DEHAZE desarrollará técnicas de eliminación de niebla que no asuman una distribución homogénea de la luz y la niebla. Creará el primer conjunto de datos de imágenes que incluye pares de imágenes reales, tanto opacas como claras y se entrenarán redes neuronales profundas para obtener imágenes claras a partir de imágenes con niebla.

Objetivo

In presence of haze, small floating particles absorb and scatter the light from its propagation direction. This results in selective and significant attenuation of the light spectrum, and causes hazy scenes to be subject to a loss of contrast and sharpness for distant objects. Besides, most computer vision and image processing algorithms (e.g. from feature extraction to objects/scene detection and recognition) usually assume that the input image is the scene radiance (haze-free image), and therefore strongly suffer from the color-shift, and low-contrast induced by hazy conditions. For instance, in normal visibility conditions the Traffic Sign Detection and Recognition (TSDR) module of the existing ADAS systems reaches a detection rate averaging around 90%, but drops below 40% in case of haze or poor illumination conditions1. Therefore, many recent works have explored inverse problem formulations and have designed dedicated image enhancement methods to address the dehazing problem. However, to estimate their key internal parameters (e.g. airlight in Koschmieder’s light transmission model), most of those solutions assume homogeneous distribution of light and haze, which is rarely the case in practice (e.g. lighting is non-uniform in space and frequency during the night, attenuation caused by haze depends on the light frequency).

Image dehazing thus remains a largely unsolved problem in case of dense and non-homogeneous haze scenes.

As a federating objective, our project aims at implementing dehazing methods that are suited to dense and non-homogeneous hazy scenes. This implies the following tasks:
(O1) build up the first (world-wide) image dataset including pairs of hazy and haze-free scenes, for which hazy scenes include real, dense, and non-homogeneous haze;
(O2) develop and train deep dehazing neural networks to derive the dehazed images from hazy inputs.
(O3) train deep image interpretation models that are suited to images captured in adverse conditions.

Coordinador

UNIVERSITE CATHOLIQUE DE LOUVAIN
Aportación neta de la UEn
€ 178 320,00
Dirección
PLACE DE L UNIVERSITE 1
1348 Louvain La Neuve
Bélgica

Ver en el mapa

Región
Région wallonne Prov. Brabant Wallon Arr. Nivelles
Tipo de actividad
Higher or Secondary Education Establishments
Enlaces
Coste total
€ 178 320,00