Parking Slot Prediction and Face Recognition Based Parked Vehicle Theft Prevention in Smart System using Deep Learning

Abstract

More than a million cars are on the roadways of a contemporary major city, but more parking spots are needed to accommodate them. Locating vacant parking places in most contemporary cities might take time, especially during busy periods like festival seasons. In the traditional parking system, drivers face considerable losses in terms of money, productivity and time which is wasted in search of parking spots in densely populated areas. Hence, it can be said that the traditional parking systems are not capable of providing a smooth parking experience to the drivers along with reducing the parking search traffic on the roads. This highlights the rationale of adopting advanced technologies to make the urban transport system modern and ease the problem faced by the drivers. This project proposes a Smart Parking System utilizing Edge Computing and Deep Learning algorithms to seamlessly link multiple parking stations into a unified network, establishing a shared parking system. To address security concerns in highly restricted areas such as residential zones, military bases, and government buildings, the system functions as a centralized automatic vehicle identifier for owner verification. Deep Learning algorithms, such as Convolutional Neural Networks used to recognize the driver/owner face of a vehicle during the departure phase, fortifying security measures and thwarting potential vehicle theft