A Survey of Computer Vision Techniques and Technologies for Interactive Media Design in Extended Reality Environments

The integration of computer vision techniques and technologies has become increasingly prevalent in designing natural and intuitive interactions, particularly in extended reality environments. This study aims to explore current trends, challenges, and practices in utilizing computer vision for interactive media design. A systematic literature review was conducted following PRISMA guidelines, based on a PICO-structured eligibility framework. From 94 reviewed studies, 11 computer vision techniques were identified in the design of interactions across virtual, augmented, and mixed realities. Emerging trends and technological dependencies were examined by considering design contexts, user interface utilization, hardware preferences, software libraries, and machine learning models. While many existing studies combine multiple computer vision techniques, this study provides a detailed categorization that clarifies their practical applications in interaction design. The study offers both a reference framework for future research and a practical guide for designers, researchers, and novice practitioners seeking to integrate computer vision into interactive media.

计算机视觉（Computer Vision）技术与方案的集成在打造自然直观的交互体验过程中愈发普及，在扩展现实（Extended Reality, XR）环境中尤为常见。本研究旨在探究利用计算机视觉开展交互媒体设计（Interactive Media Design）的当前趋势、挑战与实践路径。本研究遵循PRISMA声明，基于PICO结构化纳入排除标准框架开展了系统性文献综述（Systematic Literature Review）。本次综述共梳理94项相关研究，从中提炼出可应用于虚拟现实（Virtual Reality, VR）、增强现实（Augmented Reality, AR）及混合现实（Mixed Reality, MR）交互设计的11类计算机视觉技术。研究从设计场景、用户界面应用、硬件选型、软件库及机器学习模型（Machine Learning Models）多个维度，对新兴趋势与技术依赖关系展开了剖析。尽管现有诸多研究均融合了多种计算机视觉技术，但本研究通过细致的分类梳理，明晰了各类技术在交互设计中的实际应用场景。本研究既为后续相关研究提供了参考框架，也为希望将计算机视觉技术融入交互媒体领域的设计师、研究人员及入门从业者提供了实用指南。