Evaluation of classic colour constancy algorithms on spectrally rendered ground-truth

Authors: Toscani, M., Chen, T. and Guarnera, G.C.

Journal: I Perception

Volume: 54

Issue: 7

Pages: 478-502

eISSN: 2041-6695

DOI: 10.1177/03010066251345871

Abstract:

The limited availability of spectral images poses a significant challenge to the field of colour science. To address this issue, we spectrally rendered naturalistic images, enabling us to investigate the performance of classic colour constancy algorithms, including Grey-World, White-Patch, Grey-Edge, Shades-of-Grey, and Gamut-Mapping. We generated 4,096 physically based rendered scenes under different coloured illuminations, including a spectrally neutral illumination. We evaluated each algorithm by (1) comparing the illuminant estimated by the algorithm with the actual illuminant used for rendering and (2) assessing the performance based on the entire scene rendered under the neutral illuminant. The White-Patch algorithm consistently performed relatively well, while Gamut-Mapping emerged as the top-performing algorithm when evaluating the whole scene. However, it exhibited poor performance in estimating the ground-truth illuminant. We conducted a perceptual experiment to measure human colour constancy across a representative selection of scenes from our database using an asymmetric colour matching task. The results indicated that predictions from the algorithms that performed best when evaluated on the whole scene – white patch and gamut mapping – best approximate human performance. Indeed, the function of colour constancy is to stabilise the colour of all surfaces in a scene, rather than to estimate the colour of the illumination.

Source: Scopus

Evaluation of classic colour constancy algorithms on spectrally rendered ground-truth.

Authors: Toscani, M., Chen, T. and Guarnera, G.C.

Journal: Perception

Volume: 54

Issue: 7

Pages: 478-502

eISSN: 1468-4233

DOI: 10.1177/03010066251345871

Abstract:

The limited availability of spectral images poses a significant challenge to the field of colour science. To address this issue, we spectrally rendered naturalistic images, enabling us to investigate the performance of classic colour constancy algorithms, including Grey-World, White-Patch, Grey-Edge, Shades-of-Grey, and Gamut-Mapping. We generated 4,096 physically based rendered scenes under different coloured illuminations, including a spectrally neutral illumination. We evaluated each algorithm by (1) comparing the illuminant estimated by the algorithm with the actual illuminant used for rendering and (2) assessing the performance based on the entire scene rendered under the neutral illuminant. The White-Patch algorithm consistently performed relatively well, while Gamut-Mapping emerged as the top-performing algorithm when evaluating the whole scene. However, it exhibited poor performance in estimating the ground-truth illuminant. We conducted a perceptual experiment to measure human colour constancy across a representative selection of scenes from our database using an asymmetric colour matching task. The results indicated that predictions from the algorithms that performed best when evaluated on the whole scene - white patch and gamut mapping - best approximate human performance. Indeed, the function of colour constancy is to stabilise the colour of all surfaces in a scene, rather than to estimate the colour of the illumination.

Source: PubMed

Evaluation of classic colour constancy algorithms on spectrally rendered ground-truth

Authors: Toscani, M., Chen, T. and Guarnera, G.C.

Journal: PERCEPTION

Volume: 54

Issue: 7

Pages: 478-502

eISSN: 1468-4233

ISSN: 0301-0066

DOI: 10.1177/03010066251345871

Source: Web of Science (Lite)

Evaluation of classic colour constancy algorithms on spectrally rendered ground-truth.

Authors: Toscani, M., Chen, T. and Guarnera, G.C.

Journal: Perception

Volume: 54

Issue: 7

Pages: 478-502

eISSN: 1468-4233

ISSN: 0301-0066

DOI: 10.1177/03010066251345871

Abstract:

The limited availability of spectral images poses a significant challenge to the field of colour science. To address this issue, we spectrally rendered naturalistic images, enabling us to investigate the performance of classic colour constancy algorithms, including Grey-World, White-Patch, Grey-Edge, Shades-of-Grey, and Gamut-Mapping. We generated 4,096 physically based rendered scenes under different coloured illuminations, including a spectrally neutral illumination. We evaluated each algorithm by (1) comparing the illuminant estimated by the algorithm with the actual illuminant used for rendering and (2) assessing the performance based on the entire scene rendered under the neutral illuminant. The White-Patch algorithm consistently performed relatively well, while Gamut-Mapping emerged as the top-performing algorithm when evaluating the whole scene. However, it exhibited poor performance in estimating the ground-truth illuminant. We conducted a perceptual experiment to measure human colour constancy across a representative selection of scenes from our database using an asymmetric colour matching task. The results indicated that predictions from the algorithms that performed best when evaluated on the whole scene - white patch and gamut mapping - best approximate human performance. Indeed, the function of colour constancy is to stabilise the colour of all surfaces in a scene, rather than to estimate the colour of the illumination.

Source: Europe PubMed Central