Mahjong is widely recognized as a board game enjoyable for all ages, but its complex rules, such as score calculation, often burden players and create a high barrier to entry for beginners. This study proposes “Projection Mahjong,” a novel Mahjong interaction aimed at augmenting the real-world Mahjong game experience. By incorporating the advantages of app-based Mahjong, specifically “automatic score calculation” and “visual and audio effects,” into traditional in-person Mahjong (hereafter, traditional Mahjong), we extend the traditional real-world experience. Object detection using deep learning recognizes players’ tiles, enabling automatic score calculation and management. Projection mapping and sound effects add dynamic enhancements to traditional Mahjong. We evaluate the system via three experiments: RQ1 (tile-recognition accuracy), RQ2 (questionnaire on co-present communication and use of beyond-tabletop information), and RQ3 (score-calculation time). RQ1 achieves 99% accuracy on hand-tile detection. RQ2 finds that players perceive easier conversation and better use of nonverbal cues compared with app-based play, while preserving the feel of physical tiles. RQ3 shows faster score computation: manual scoring by proficient players takes about twice as long as with Projection Mahjong. Together, these results support projection-based, on-surface augmentation as a practical path to lowering rule/scoring burden without sacrificing the social dynamics of traditional Mahjong. A demonstration video is available at https://youtu.be/LsOJOzUNDQY .
Kato et al. (Fri,) studied this question.