Off the Planckian Locus: Using 2D Chromaticity to Improve In-Camera Color

2025-11-21

AI Summary

Traditional camera color mapping relies on one-dimensional correlated color temperature (CCT) interpolation, optimized for Planckian sources (e.g., illuminants A and D65), and thus fails to accurately characterize modern non-Planckian LED sources deviating from the blackbody locus—leading to color reproduction distortion. This paper proposes a lightweight multilayer perceptron (MLP)-based color mapping method operating directly in the two-dimensional CIE 1976 (u′, v′) chromaticity space, abandoning CCT-based parametric modeling to enable fine-grained representation of non-blackbody spectra. The model is calibrated using a standard light source box and trained on representative LED spectral data, ensuring compatibility with both multi-illuminant scenes and conventional Planckian sources. Experiments demonstrate a 22% average reduction in color difference angle across diverse LED illuminants, with low computational overhead suitable for real-time in-camera processing. The core contribution is the first integration of (u′, v′) chromaticity features with a lightweight MLP for camera color mapping, significantly enhancing color fidelity under non-Planckian illumination.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Traditional in-camera colorimetric mapping relies on correlated color temperature (CCT)-based interpolation between pre-calibrated transforms optimized for Planckian illuminants such as CIE A and D65. However, modern lighting technologies such as LEDs can deviate substantially from the Planckian locus, exposing the limitations of relying on conventional one-dimensional CCT for illumination characterization. This paper demonstrates that transitioning from 1D CCT (on the Planckian locus) to a 2D chromaticity space (off the Planckian locus) improves colorimetric accuracy across various mapping approaches. In addition, we replace conventional CCT interpolation with a lightweight multi-layer perceptron (MLP) that leverages 2D chromaticity features for robust colorimetric mapping under non-Planckian illuminants. A lightbox-based calibration procedure incorporating representative LED sources is used to train our MLP. Validated across diverse LED lighting, our method reduces angular reproduction error by 22% on average in LED-lit scenes, maintains backward compatibility with traditional illuminants, accommodates multi-illuminant scenes, and supports real-time in-camera deployment with negligible additional computational cost.

Problem

• Improving color accuracy for non-Planckian LED lighting
• Replacing 1D CCT interpolation with 2D chromaticity mapping
• Enabling robust colorimetric mapping under diverse illuminants

Innovation

• Transitioning from 1D CCT to 2D chromaticity space
• Using lightweight MLP for robust colorimetric mapping
• Training MLP with LED sources for improved accuracy