Depth-Enhanced Object Identification and Interaction in Augmented Reality Using an ARKit-Based Framework

Abstract

Augmented Reality (AR) has emerged as a potent technology for instantaneous object recognition and interactive representation. This research introduces an augmented reality identification and interaction framework created using Apple’s ARKit platform. The experimental implementation uses scene data and object annotations sourced from GitHub repositories related to ARKitScenes. The dataset comprises geographical mapping data, RGB image frames, depth maps, 3D scene reconstruction information, camera intrinsics, object bounding boxes, and pose estimation parameters. The system considers essential characteristics such as camera focal length, intrinsic matrix values, depth resolution, object coordinates (x, y, z), and feature descriptors derived from image frames. The proposed method incorporates feature extraction, spatial alignment, and real-time interaction components to improve object identification accuracy in augmented environments. The performance assessment indicates consistent object identification, stable tracking, and effective interaction management across diverse lighting and scene conditions. The framework emphasizes the practical capabilities of ARKit-based systems for immersive identification and interaction applications in smart environments, education, and industrial support systems.