Multi-View Human Activity Analysis Framework

Bys: Vikrant Sharma, Ketanpreet Kaur, Gagandeep Singh

This book focuses on the design and development of an advanced Smart Video Surveillance System (SVSS) to address growing global security concerns. With the rapid increase in thefts and public safety issues, traditional surveillance systems are no longer sufficient, as they mainly perform basic tasks like face detection, object detection, and counting. The book highlights the need for deeper analysis of human activities in video data to detect suspicious behavior more effectively. The core contribution of the work is the proposal of a Human Activity Analysis Framework integrated with Fog Computing. This framework aims to process video data closer to the source (at the fog layer), enabling faster and more efficient real-time analysis. To achieve accurate human activity recognition, the authors introduce a hybrid deep learning model combining Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks. This model captures both spatial (image-based) and temporal (sequence-based) features from video frames, significantly improving action recognition performance.

• Format: Paperback
• Language: English
• ISBN: 620989481X
• ISBN13: 9786209894817
• Release Date: April 2026
• Publisher: LAP Lambert Academic Publishing
• Length: 208 Pages
• Weight: 0.63 lbs.
• Dimensions: 9.0" x 0.5" x 6.0"