Development and implementation of face recognition system using neural networks

Chang, Kai Xin (2021) Development and implementation of face recognition system using neural networks. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

	Text (24 Pages) Development and implementation of face recognition system using neural networks.pdf - Submitted Version Download (1MB)
	Text (Full Text) Development and implementation of face recognition system using neural networks.pdf - Submitted Version Restricted to Repository staff only Download (3MB)

Abstract

One of the most widely used technologies in the world today is the facial recognition. It is used in biometric security systems to identify a person digitally before granting the access to the system or the data in it. The news of a female being kidnapped and murdered by a police officer in London has become a global topic and many are concerned about the safety of women in daily life. The researcher’s motive is to help females and children under threat so that they can be rescued before it is too late. The objectives of this project are, to develop a device that will capture the image of a kidnapper as evidence for future reference and send the captured image to the family of the victim through email, to design a face recognition system to be used in searching kidnap suspects and to determine the best training parameters for the convolution neural network layers used by the proposed face recognition system. The system is divided into two parts, the hardware, and the software, where the hardware part consists of the ESP32-CAM programmed by Arduino IDE, which can capture image of the kidnapper and send it to the email of the victim’s family through SMTP server; and the software which is a face recognition system built in MATLAB to match the image captured by the hardware with the faces stored in a database that resembles the database of the authorities. The system is tested with different images captured by the hardware and the software is able to recognise and compare the face with the image database and lastly provide the name or identity of the person. The hardware device of the system is small and compact to be carried around with ease, it can be attached to random items so that the user can easily trigger the button on the device to capture the image of the other person when they are in danger. The best training parameters for the proposed CNN model are kernel size of 5x5, 32 and 64 filters for first and second convolutional layers and learning rate of 0.001. The proposed system is robust as its overall face recognition accuracy is 98.48%.

Actions (login required)

View Item