Multiplatform mobile augmented reality for markerless visualization of urban environmental and green city concepts using 3D trees

Abstract

This study developed a markerless Augmented Reality (AR) application for visualizing dynamic 3D tree growth to support Green City planning and environmental education. The system simulates the development of five urban tree species based on morphological parameters, including height, trunk diameter, and crown width, using ecological datasets processed through polynomial interpolation. The application was built using Unity AR Foundation and operates on both ARCore and ARKit platforms, achieving stable outdoor visualization with response times of 140–245 ms and ecological simulation errors below 8%. A multiplatform implementation is provided to support smartphones, tablets, and wearable devices. System validation was conducted using black-box testing to evaluate functional correctness, system behavior, and interaction outcomes without examining internal implementation details. The testing process focused on core features such as plane detection, object placement, tree growth visualization, and species selection. The results show that all primary functionalities perform as expected across multiple device types, with stable operation and no critical functional errors identified during testing. The developed application supports urban planners in evaluating canopy expansion, shading effects, and spatial interactions between trees and urban infrastructure, while providing educators with an interactive tool for teaching Green City concepts, urban ecology, and long-term tree growth dynamics. Future work will focus on improving plane detection under low-light conditions, optimizing rendering performance across heterogeneous mobile devices, and incorporating advanced sensing technologies to enhance tracking stability and deployment reliability. Unlike existing AR-based environmental visualization systems that primarily present static 3D objects, the proposed system integrates ecological growth parameters into markerless AR to simulate dynamic tree development and long-term canopy expansion in real-world urban environments.