Keyless access via facial recognition

Thesis (Bachelor of Science in Computer Engineering) -- University of Rizal System-Morong.

EXECUTIVE SUMMARY: This study aimed to develop a Keyless Access via Facial Recognition to enhance the security of a vehicle such as motorcycle from being stole with the use of portable and cheap device. The study aimed to develop a prototype of Keyless Access via Facial Recognition using locally available materials. The researchers used Developmental Method and Descriptive Method along the study. This means that this study is based on systematics of designing, developing, and evaluating instructional programs, processes and products that must meet criteria of internal consistency and effectiveness. This study was conducted at the University of Rizal System-Morong Campus during the school year 2016-2017. And the respondents are selected using purposive sampling technique and they are all using vehicles or have knowledge in using vehicle. The respondents are categorized as chosen students and professors of the University and experts who have lot of knowledge on vehicles. The researchers utilized and adopted questionnaire checklist to evaluate the performance of the system in terms of accuracy, durability, functionality, portability, safety and maintenance to ascertain the acceptability of the proposed system. The specification data of the software (python, OpenCV, Raspbian Operating System) hardware (Raspeberry Pi, PiCamera, Wi-Fi dongle, Boost Converter, Relay mpdule) are needed in designing the entire device. During the design, proper, the researchers conducted and developed the device after the construction, initial testing and series of trial was made to check if there are some adjustments and conditions needed to complete the project device. This study is limited in a way that the device can only make accurate recognition when the background or lightning where the system was trained is the same as the environment where it is taking samples from. Moving or static background can affect the system's confidence in recognition. The researchers developed this device to lessen the cases of stolen vehicles and to identify the intended owner. The summary of design result shows the level of acceptability in terms of accuracy, durability, functionality, portability, safety and maintenance is "Highly Acceptable" as gleaned by the student, professors and experts.