PIR Sensor-Based AAL Tool for Human Movement Detection: Modified MCP-Based Dictionary Learning Approach

Abstract

Development of sensor(s)-based, signal processing-aided, low-cost ambient-assisted living tools (AALs), essentially for assisting elderly people, home automation, and remote monitoring purposes, has become an important research domain. Within this domain, developing intelligent systems for human movement recognition in specific directions has become a very important problem statement. This article shows how a sophisticated, low-cost, integrated system can be developed using an indigenously developed hardware–software combine. The solution employs around four pyroelectric infrared (PIR) sensor based hardware systems coupled with a novel dictionary learning algorithm. The work successfully carries out the recently proposed multiple cluster pursuit (MCP)-algorithm-based dictionary learning for the human detection problem and then proposes a new variant of MCP algorithm, called modified MCP algorithm, for this purpose. Extensive real-life performance evaluations have been performed to demonstrate the suitability of MCP and the modified MCP algorithms for the problem under consideration.