The machine learning techniques paired with augmented reality (AR) systems extend the capabilities of surveillance systems for the social good through analyzing and extracting complex patterns in a simple and cost-eff...
详细信息
ISBN:
(纸本)9798400716218
The machine learning techniques paired with augmented reality (AR) systems extend the capabilities of surveillance systems for the social good through analyzing and extracting complex patterns in a simple and cost-effective AR-based app running on any mobile device. The potential of this app in crime prevention provides opportunities for this study to investigate the development of a portable, inexpensive surveillance system using an Android device that mainly focuses on detecting deadly weapons to promote public safety, especially in urbanized areas. The detection of deadly weapons beforehand is a proactive act before the crime happens. The study utilized the detection models offered by tensorflow, an open-source machine-learning platform with a broad and adaptable ecosystem of tools and libraries. After the functional and performance tests, the study found that the AR-based weapon detection system could detect the weapons and recognize and classify them based on the computation of the detection confidence. Thus, the cost-efficient and portable system showcased its potential as a proactive surveillance tool that detects and alerts for the weapon's existence. However, there is still a need to improve its detection capability for it to be able to increase its detection rate, especially in crowded areas, at far distances, in dark places, and those obstructed and concealed weapons.
暂无评论