A Mobile Augmented Reality System Recognizes Three Image Targets and Preserves Virtual Object Placement

Abstract

Children naturally learn and play by moving through physical spaces, manipulating objects, and building stories around what they see. Most digital experiences, however, keep children focused on a screen and disconnected from their surroundings. Augmented reality (AR) bridges this gap by transforming familiar physical objects into interactive parts of a digital world. A convincing AR experience must accurately recognize multiple physical targets, attach the correct digital content to each, preserve its position and orientation, and transition smoothly as the child moves between targets. Failures in these areas can interrupt the experience, breaking the sense of exploration. To address this challenge, I developed a mobile, image-target-based AR module for AnchorPlay, a child-centered interactive system connecting physical play with digital storytelling. Created using Unity and Vuforia, the module includes three physical image-targets: a house, a tree, and a pond. Each target serves as an entry point into the experience by activating a corresponding three-dimensional virtual object in the child's physical environment. During functional testing, the system successfully recognized all image targets (3/3) and displayed the correct virtual content for each one. It also supported transitions among the house, tree, and pond without disrupting the interaction sequence or changing the objects’ placement and orientation. My current direction focuses on completing the interaction flow and refining how children discover and engage with the virtual content. By addressing challenges of multi-target recognition, mobile deployment, and continuous interaction, this project lays the foundation for an AR experience that transforms physical spaces into interactive digital worlds.