Implementation of Enhanced Axis Aligned Bounding Box for Object Collision Detection in Distributed Virtual Environment

Abstract

Axis Aligned Bounding Box (AABB) is a popular method for object collision detection in distributed virtual environment (DVE). Actually, besides AABB several researchers have found better methods but those methods are not used in DVE. The simplicity of AABB is one of reasons. Another side, researchers have found several methods to make a scalable DVE to accommodate huge users. Object-based simulators architecture is one of them. It uses a simulator for an object in DVE. This paper aims to determine the comprehensive effects of implementing AABB in DVE and to enhance AABB implementation in DVE using object-based simulators architecture in order to be able to use other object collision detection methods in DVE. This research had four stages. First stage, we developed DVE application. The second stage was running AABB in DVE. In this stage, we also determined characteristics of AABB implemented in DVE. Third, we implemented AABB in object-based simulators architecture. This architecture used more simulators to manage objects. Finally, we analyzed and compared the results to the common existing DVE. There were three parameters measured in this research: runtime, frame rate and CPU usage of simulation application. The experiment results showed that computation complexity was the most important thing to be considered in DVE because DVE always updates its environment. Adding a few workloads into AABB degraded strictly performance of DVE. The results also showed that the enhanced AABB implementation in DVE using the object-based simulators architecture could rich the user experiences. It was achieved by the smaller runtime, higher frame rates and smaller CPU usages than the common DVE. This novel approach could be used to implement other better collision detection algorithms in DVE such as oriented bounding box (OBB).