Parking Lots Detection in Static Image Using Support Vector Machine Based on Genetic Algorithm

Abstract

The increasing ownership of four-wheeled vehicles creates a new problem: difficulty finding available parking spaces in significant places. Issues citizens often experience these in the city centre who use their cars for transportation. This causes the detection of parking lots to attract the attention of researchers. Unfortunately, this system fails when the vehicle occupies more than one place or when the parking lot has a different parking space. In this study, an automation method is proposed that utilises a combination of gray level co-occurrence matrix (GLCM) and support vector machine (SVM) methods with genetic algorithm (GA) optimisation techniques or called SVMGA. The results showed that the SVMGA could provide an accuracy performance of 96.99% for training data and 94.36% for test data. The results of this study are expected to help the queue of people to find a parking space and reduce the density of lines in the parking lot.