Real-time hand gesture recognition system based on Associative Processors

In this paper, a general-purpose similarity-measure recognition system using associative processor (AP) chips for real time hand gesture recognition is proposed. In order to get hand gesture feature vectors, the system adopts a vision-based hand tracking approach by using hand gesture segmentation algorithm. The system downloads those feature vectors data from large hand gesture feature vectors data base into the on-chip cache memory of an AP, then performs gestures matching in an extremely short time. Although gestures recognition processing is computationally very expensive by software, latency free recognition becomes possible due to the highly parallel maximum-likelihood matching architecture of the AP chip. In this study, we propose a solution about hand gestures recognition using large testing/training data and the hardware-accelerated matching architecture for human-computer interaction. Using a prototype AP chip implemented in field-programmable gate arrays (FPGA), the effectiveness of such application systems has been demonstrated.