Design team members: Sameer Rehman, Simon So, Clayton Tso
Supervisor: Professor Paul Fieguth
Background
As it exists today, online shoppers are limited to visualizing their purchases through static images that are photographed in environments that rarely imitate real life conditions. The inability of a customer to completely visualize the product significantly reduces their incentive to purchase it. In cases where videos of the merchandise are made available, customers are still restricted to viewing the product at angles and conditions that the seller has selected to display. There have been attempts to give the consumer more control, such as with panoramic imaging and virtual reality modelling, with behavioural studies showing the latter to be very successful. The goal of this project is to take this one step further and provide a more immersive online shopping experience, using augmented reality (AR) to allow the user to easily visualize 3-D products within their own environments. Online merchants can benefit from this technology through improved consumer product knowledge, brand attitude, and purchasing intention.
Project description
Augmented reality allows a whole new level of interaction between the user and product, modifying our perception of reality in real-time using computer generated virtual objects. There are existing augmented reality systems with a variety of different applications, each with key differences and shortcomings compared to our proposed design. The Tinmith AR system and MARS (Mobile Augmented Reality System) from Columbia University allow the user to enjoy an augmented outdoor environment but requires them to carry a bulky computer unit and headset. Michigan State University has developed an augmented reality shopping system using a portable LCD that overlays 2D data such as price comparisons, promotions and advertisements: Our proposed design shall take this one step further and enable any individual with the applicable hardware and Internet access to view product models in 3-D, even if they possess low technical aptitude. This enhances the shopping experience by allowing users to view customized products at any angle within the context of their preferred environment, and without having to physically visit a retail store.
Design methodology
An existing open source AR development platform (Augmented Reality Toolkit - ARToolkit in short) based on C and OpenGL was found to perform many of the basic functions thus specified. It already provides a driver for inputting a video feed from any Windows-compatible USB webcam as well as a basic framework with sample markers upon which simple 3D models can be overlaid. Rather than reinvent the wheel, this existing platform was leveraged, its limitations identified and an improved implementation designed to achieve an even more complete immersive shopping solution. The focus will be on finding solutions to ARToolkit's limitations such as the ability to detect only single markers, problems with occlusion, no "out-of-sight" tracking ability, etc.