AR Powered Interior Decorator: A Real-Time Augmented Reality Application for Interior Decoration and Furniture Visualization

Abstract

The rapid proliferation of smart devices and e-commerce has transformed how consumers select and purchase furniture. Traditional interior planning relies on imagination, 2D catalogs, and manual measurements, frequently resulting in incorrect purchasing decisions and poor space utilization. This paper presents the AR Powered Interior Decorator, a real-time Android application that leverages Google ARCore's SLAM (Simultaneous Localization and Mapping) for environment understanding, Sceneform/Filament for physically- based 3D rendering, and AABB (Axis-Aligned Bounding Box) collision detection for placement validation. The system enables users to select GLB furniture models from scalable datasets - SketchFab (46,000+ models) and CGTrader (51,300+ assets) - and place them within their actual room via smartphone camera. Gesture- based interaction supports real-time object translation, rotation, and scaling. Multi-object placement allows complete room layout experimentation. Evaluation across multiple room types confirms stable SLAM tracking (< 2 cm drift over 5 minutes), consistent rendering at 55–60 fps on mid-range devices, and 100% collision rejection accuracy. The proposed system demonstrates a practical, scalable approach to AR-driven furniture visualization that meaningfully reduces purchase uncertainty and improves interior planning outcomes.