DefinePK

Abstract

Ensuring safety of the pilgrims is the top priority of Saudi government. More than two millions of pilgrims visit the holy cities of Makkah and Madinah, in KSA, to perform Hajj and Umrah rituals every year. Therefore, kingdom considers health, safety and security a red line that is not supposed to be crossed at any extent, considering the magnitude of related consequences. A higher tendency of persons is found who get separate from their group of members during Hajj and Umrah rituals; especially in the case of children and old-aged ones. Our survey from pilgrims who lost their elders or children shows that amongst the missing 8% were elderly persons and 92% were children. The lost elderly persons and children need an efficient system to track them. Therefore, this paper proposes a system model and a prototype application for effective reporting and tracking of lost and found persons in a huge crowd gathering scenario of Al-Nabawi mosque Madinah. Our proposed system’s model aims to facilitate both pilgrims and the center of lost and found persons in Madinah. It allows pilgrims to report and search lost persons effectively. It matches the reported spatio-temporal contextual information of lost persons along with integrated support of the face recognition and biometric verification techniques. We used an existing deep learning method based on a face recognition algorithm which results in high accuracy during small-scale testing on labeled faces in the Wild benchmark. The tracking procedure is made more efficient by coupling the face recognition technique employed with biometric verification and reported spatio-temporal features of lost pilgrims.