Title: Windowing-based threshold technique to play the simple breakout game at neutral attention level
Authors: Gauttam Jangir; Rakesh Kumar
Addresses: Department of Computer Science and Engineering, National Institute of Technical Teachers Training and Research, Chandigarh, 160019, India ' Department of Computer Science and Engineering, National Institute of Technical Teachers Training and Research, Chandigarh, 160019, India
Abstract: Brain computer interface (BCI) applications focus on an assistive care and provide an alternative communication. In the proposed work, the famous breakout game implementation is done to evaluate the players' EEG signals (low alpha, high alpha, low beta and high beta) for the constant threshold and windowing-based threshold (WBT) along with its variations. WBT variations like sliding windowing-based threshold (SWBT) and conditional sliding windowing-based threshold (CSWBT) are used to provide the support during game playing. The proposed WBT algorithm is a process of grouping stream data and applying operations on them. EEG signals were evaluated by using ANOVA method, after passing through band pass filter. Common average reference (CAR) was also applied to reduce EEG artefacts. WBT algorithm results have been compared with existing constant threshold. Performance gain is 70% in terms of accuracy which shows that WBT helps to play, the breakout game at neutral attention level.
Keywords: brain computer interface; windowing-based threshold; WBT; EEG; visual evoked potential.
International Journal of System of Systems Engineering, 2017 Vol.8 No.2, pp.147 - 173
Received: 30 Jan 2017
Accepted: 03 Sep 2017
Published online: 01 Dec 2017 *