KLASIFIKASI KEPADATAN LALU LINTAS PADA SIMPANG BERSINYAL MENGGUNAKAN FUZZY LEARNING VECTOR QUANTIZATION; CLASSIFICATION OF THE TRAFFIC DENSITY AT SIGNALIZED INTERSECTION USING FUZZY LEARNING VECTOR QUANTIZATION

UTOMO, PRADITYO, Agus Harjoko

2016 | Disertasi | FMIPA

Population growth in Indonesia have caused traffic congestion, especially at signalized intersections. To overcome those congestions, there have been some research on adaptive traffic lights. In this study, a system to classify the traffic density at the intersection is developed, where the classification of the traffic density is measured based on the area of vehicle queues. For that, digital image processing methods and Fuzzy Learning Vector quantization (FLVQ) are used. Digital image processing methods which is used to calculate the area of the vehicles queues consist of grayscale, threshold, edge detection, Hough Transform, image elimination, image subtraction, erosion, dilation, and connected component labeling. Having obtained the area of vehicle queues, traffic density classification is done using FLVQ algorithm. The system obtained accuray of 93.75% for the traffic density test. In addition, in the recognition test, it obtained 85,56% accuracy when using erosion and 88,89% accuracy without erosion.

Kata Kunci : Traffic Density, Classification of The Traffic Density, Fuzzy Learning Vector Quantization, Hough Transform, Connected Component Labeling