Date of Award


Document Type


Degree Name

Master of Science (MS)


College of Science and Mathematics


Computer Science

Thesis Sponsor/Dissertation Chair/Project Chair

Jiacheng Shang

Committee Member

Bharath K. Samanthula

Committee Member

Li Dawei


For in-home security, intelligent operations like top individual recognition and minimizing losses due to home break-ins, emergencies, and fraud are keys to success. This application integrates the closed-circuit television (CCTV) camera and the deep learning algorithms used to process these images. Automated intrusion detection alerts, real-time fire alerts, smart checkout, and potentially fraudulent point of sale (POS) transactions are its main features. Dynamic intrusion with machine learning is a software program in which the price of certain products changes over time through an algorithm that considers a variety of pricing variables. The face locator is a part of the algorithm that locates and detects motion by using the image search function. The system collects all available product locations from the live videos from multiple cameras. This is a helpful feature for finding misplaced products and detecting POS user fraud. This intrusion detection system (IDS) records POS transaction details on the screen as an overlay on video images to reduce home break-ins. To improve the ease and speed of transaction searches, the faces of individuals are used to search for disputed cases. Smart Checkout System (SCS) utilizes a self-service kiosk where users can generate bills by showing products to the linked camera. SCS uses Google vision technology to identify products. Motion detector and queue detection will detect long queues at the checkout counter in real-time and open new lanes to speed up the transaction, improve the experience, and reduce the number of abandoned purchases. Face recognition premium and alerts can also be provided.

File Format