IoT application: 360 Smart Parking Guidance System by Using Image Processing

Recently, a limited parking space in a city area becoming a very serious issue since the world population keep growing which then cause an increase in number of car drivers. Some big companies like shopping malls and other public facilities already deployed their smart parking system on each building for their customer conveniences. However, there are still a lot of building that do not own it because the system required a lot of investment. Larger parking area need a very high initial cost especially for installation of the sensors to each available parking lot. The proposed Smart Parking Guidance System in this research were depending on a 360-camera integrated with an image and video processing technology. This framework can detect and notify users regarding parking space availability and it is built specifically for a closed parking space area. A single 360-camera may replace several ultrasonic sensors that implemented on a conventional smart parking system. For this research, the 360-camera simulated using 2 VGA camera placed at a specific angle to capture the presence of a car. Later, video captured were processed by using Haar-cascade training method. Based on the result, the integrated framework working as expected. However, several performance issues need to be considered. The high performance Haar-cascade file demand for high-end CPU for a smooth image processing to happen. Other than that, a very good quality camera with special feature is crucial to ensure the detection can be done accurately. Further details are discussed and met the requirements.