SUMMARY
The development of smart city concept, particularly in smart parking systems, has not solved a problem that occurs in metropolitan areas, such as in urban areas where the population has continued to rise, resulting in high demand for private vehicles and parking spaces. Finding a parking space is known as the most common issue the drivers have had specifically on peak hours’ time. During peak hours, the difficulty arises as many people look around to find vacant parking space at once, which causes many negative impacts on cities and drivers themselves, such as pollution, traffic congestion, traffic accidents, waste of time and fuel, emotions and so on. As a solution, smart parking system exist to equip parking lots with many different types of sensors to automatically detect free parking space that would guide drivers to find the nearest car parking space as efficient as possible. An effective smart parking system can solve this problem and make better use of parking resources. However, many smart parking systems still uses embedded sensors that are expensive for installation and inefficient. This paper presents a review of the existing approaches to the smart parking system. This paper focuses on a machine-vision-based technology used for smart parking system and highlights its main features, advantages and disadvantages.