An interactive floor for shape-based interactions using a client-server architecture

Interaction with large surfaces such as walls and floors has become an interesting topic among scholars and researchers. Several approaches to implement such surfaces have been explored, as well as ways of evaluating its user interaction. The objective of this paper is three-folded: first, it sought to shed light on how to implement a low-cost computer vision interactive floor, based on a shape-based interaction approach (conversely, to the common point-based interaction approach used in other solutions); second, the performance of the client-server system was studied; third, a usability study was set up and examined quantitatively the users' satisfaction and realism of their interaction with the system. The results demonstrated that a maximum of four people can interact simultaneously compromising the CPU usage up to 50% on the server and up to 90% on the client. The network traffic was also analyzed showing that due to the simplifications of the shapes sent from the server to the client, the traffic was between 127 Kbps and 759 Kbps. Moreover, the user experience and realism are enhanced using the approach proposed. Conclusions and perspectives for further development and future work are presented at the end of the article.