Recognition of Traffic Weight Using Sobel Edge Detection Method and K-Nearest Neighbor Algorithm

Eric John  G. Emberda,; Lovie Mae  N. Dalagan,

Recognition of Traffic Weight Using Sobel Edge Detection Method and K-Nearest Neighbor Algorithm

Eric John G. Emberda | Lovie Mae N. Dalagan

Abstract:

This study explored the use of Sobel Edge Detection and K-Nearest Neighbor algorithm in classifying the traffic weight of a given captured image. A software application was created that accepts as input, a snapshot of a given intersection. The application could determine the traffic weight of the given snapshot, as whether it is light, moderate, or heavy by comparing it to a database of images using the K-Nearest Neighbor algorithm. The accuracy of the result was highly dependent on the training data and the quality of the snapshot. Overall, the use of Sobel Edge Detection and K-Nearest Neighbor algorithm gave significant results in recognizing the weight of a given snapshot of traffic.