Counting the Number of Moving Vehicles by Its Type Based on Computer Vision
Yuslena Sari, Andreyan Rizky Baskara, Khair Zuniar Rahman, Puguh Budi Prakoso · 2021 4th International Conference of Computer and Informatics Engineering (IC2IE) · 2021
Traffic jam, lack of adequate information of the traffic low used for development of road infrastructure to reduce traffic jam, and possibility of human error in the execution of a heavy traffic survey become the primary background of this study. By using information technology, open up an opportunity for the development of monitoring system computer vision based. This study will calculate the total amount of moving vehicles based on its type with computer vision based with staged: ROI selection, image segmentation with Gaussian Mixture Model method, filtering process, blob detection and tracking, and vehicles classification with Fuzzy Clustering Means. Implementation of an application using visual studio 2010. The output comprises result of classification and total amount vehicles based on its type. The test application divided into a few test scenarios, namely test 1, test 2, test 3 and test 4. The accuracy obtained on each test are 36.27%, 50.47%, 60.75%, and 67.00% respectively.