Integrated Approach to Home Security and Safety Systems

This book provides an integrated solution for security and safety in the home, using Face Recognition, Voice Recognition, and Similarity Index to recognize individuals and monitor health parameters. A hybrid algorithm named PICA combines PCA and ICA, while MFRASTA combines MFCC and RASTA-PLP for voice recognition. A Similarity Index provides trust-based security. A fingertip-based application for health monitoring is also proposed, using iPhone 6s camera and 3G Dongle to send text messages in emergencies.

Format: Hardback
Length: 176 pages
Publication date: 15 October 2021
Publisher: Taylor & Francis Ltd

This comprehensive book offers an all-encompassing solution for home security and safety, encompassing both assistance in health monitoring and protection against unwanted intruders with harmful intentions. It establishes a framework that employs three distinct modules to recognize individuals at the door: Face Recognition, Voice Recognition, and Similarity Index. These three modules are combined to calculate a percentage likelihood of whether the person is known or unknown. Moreover, the system continuously monitors vital health parameters of vulnerable individuals living alone at home, enabling them to promptly call for help during emergencies.

The authors have extensively analyzed various existing biometric techniques to enhance security for individuals residing alone at home. These techniques have undergone rigorous testing using MATLAB® image processing and signal processing toolboxes, resulting in calculations based on recognition rates. A significant breakthrough in securing an individual's home is a hybrid algorithm proposed by the author, known as PICA. This innovative approach combines the strengths of Principle Component Analysis (PCA) and Independent Component Analysis (ICA), delivering superior recognition performance compared to either system used individually.

Furthermore, the authors have introduced a second proposed hybrid algorithm for voice recognition, named MFRASTA. This algorithm amalgamates features from Mel Frequency Cepstral Coefficient (MFCC) and Relative SpecTrA-Perceptual Linear Prediction (RASTA-PLP) algorithms, leading to improved recognition accuracy in voice recognition tasks. Following extensive experimentation, results have been collected based on recognition rates.

Additionally, the authors propose a third technique, known as Similarity Index, which provides trust-based security for individuals. This technique is text-independent, relying on pronunciation to recognize a person. By utilizing these three modules together, the book offers a holistic and robust solution for home security, ensuring the well-being and safety of individuals within their homes.

Weight: 436g
Dimension: 160 x 241 x 19 (mm)
ISBN-13: 9780367638412