Image color-based preset light matching algorithm for an electric vitrine

Abstract

Recent advanced light systems offer light presets to enable users to navigate a proper light. Often, the presets are labeled with target ambiance or mood, yet ambiguous for users to predict the light. This study proposes an algorithm that matches the light presets based on the image color characteristics derived from a photograph taken by users. In particular, we developed an automatic match of a light preset for an electric vitrine in which 15 RGB LEDs were linearly arrayed beneath the cover. We conducted a creativity workshop with eight light designers and composed a pool of 22 light presets for the electric vitrine. The presets attempted to cover five standard illuminants, eleven chromatic lights, and six kinds of the color spectrum. The algorithm enables users to receive the optimally matched light presets in order of hue similarities between displayed objects and the light preset. Based on stakeholders’ feedback, the algorithm UX is designed to provide three alternatives made up of the best chromatic light color, the best standard illuminant, and the best color spectrum. We expect the algorithm to be applied to different contexts, where light needs to be optimally tuned by being aware of the context characteristics.