New Delhi: Researchers at National Institute of Technology (NIT), Rourkela have secured a patent for a fully automated system which recognises intruders based on their walking pattern to detect unauthorised access in large and complex building environments, according to officials.
The "unauthorised person detection using thermal imaging and gait recognition for intra-building security" system introduces a next-generation solution to address the challenge of monitoring unauthorised access in large and complex building environments, they said.
According to Samit Ari, Professor, Department of Electronics and Communication Engineering, conventional CCTV camera-based surveillance systems require extensive manual monitoring and analysis, making them ineffective and prone to human errors.
"In large setups, tracking individuals across multiple cameras is difficult, particularly under varying lighting conditions and occlusions. To address these limitations, we developed a fully automated, non-intrusive system that can detect, identify, and track individuals using thermal imaging technology," Ari said.
"The innovation is designed to monitor suspicious activities. The integrated thermal imaging technology helps distinguish humans from the background due to lower infrared noise and enables accurate detection even in inadequate lighting conditions," he added.
To achieve this, the research team used human gait as a biometric identifier, which recognises individuals based on their unique walking pattern, thus reducing the need for manual monitoring.
"When a person enters a restricted area, the system compares their walking pattern with those of authorised individuals. If no match is found, it flags the individual as suspicious and alerts security. We developed a working prototype with three thermal cameras connected to a central server via USB interfaces.
"The system detects unauthorised individuals at entry, tracks their movement across multiple checkpoints, maintains a temporary database for unknown individuals, and determines entry and exit patterns based on movement direction. If an individual exits the restricted area, their temporary record is cleared, while historical data remains stored for future reference and forensic use," Ari said.
Developed at an estimated cost of Rs 1.90 lakh, the system uses thermal imaging for low-light and night-time operation, along with gait-based identification for non-intrusive biometric recognition. Automated multi-camera tracking enables seamless monitoring across locations, supported by centralised data processing for logging and database management.
"The system has wide-ranging applications across academic institutions and corporate campuses, as well as in defence and high-security setups where reliable identification is critical. It is also suited for deployment in industrial and research and development facilities to enhance monitoring and safety. Additionally, its capability to operate effectively in night-time and low-visibility conditions makes it ideal for surveillance operations in challenging environments.
"The developed system can be utilised to minimise dependence on manual surveillance, thereby improving operational efficiency. It supports faster and more accurate threat detection, allowing timely responses to potential risks. Designed to perform reliably in challenging conditions such as low visibility and complex environments, it also enhances forensic investigations through the availability of stored gait data for analysis and identification," he said.
This report was published from a syndicated wire feed. Apart from the headline, the EdexLive Desk has not edited the copy.