Color constancy using CNNs

Abstract

In this work we describe a Convolutional Neural Network (CNN) to accurately predict the scene illumination. Taking image patches as input, the CNN works in the spatial domain without using hand-crafted features that are employed by most previous methods. The network consists of one convolutional layer with max pooling, one fully connected layer and three output nodes. Within the network structure, feature learning and regression are integrated into one optimization process, which leads to a more effective model for estimating scene illumination. This approach achieves state-of-the-art performance on a standard dataset of RAW images. Preliminary experiments on images with spatially varying illumination demonstrate the stability of the local illuminant estimation ability of our CNN.

Illuminant estimation in multispectral imaging

“…[60,61]. Bianco et al [62] used a convolutional neural network for illuminant estimation in raw images. For generation of ground-truth illumination, shades of gray, gray edge, and gamut mapping are applied on the training data in their proposed method.…”

The Reproduction Angular Error for Evaluating the Performance of Illuminant Estimation Algorithms

“…of commercial cameras. Here we do not have access to the whole set of algorithms used in the original study [14] by Gijsenij et al (indeed, the performances supplied there for the datasets were contributed by many authors (including the recent methods [27], [28], [29], [30]) i.e. there is not a complete code repository).…”

Análise experimental de algoritmos de constância de cor e segmentação para detecção de mudas de plantas

“…Entretanto apesar da grande quantidade de algoritmos propostos para constância de cor, nenhuma solução única foi identificada, as possibilidades de aplicações e de condições ambientais são tão amplas e diversas que impedem a obtenção de uma solução universal, de forma que as abordagens estão muito mais orientadas às suas aplicações e à problemas particulares (AGARWAL et al, 2006;GIJSENIJ;Van De Weijer, 2011). Recentes avanços tem mostrado resultados promissores com o uso de redes neurais convolucionais e deep learning (BIANCO; CUSANO; SCHETTINI, 2015;LOU et al, 2015;OH;KIM, 2017), mas sua implementação requer dados de treinamento e custos computacionais relativamente altos, restringindo sua aplicação.…”