Performance Analysis of Different Spatial Domain Methods for Traffic Control Using Image Processing: A LabVIEW Approach

Abstract

In this paper, we present a scheme for performance analysis of spatial domain methods, namely, Laplacian, Arbitrary, Sobel and Prewitt operator methods for traffic control using an image processing with LabVIEW approach, including timing constraints are used to control the signal along the crossroad signal posts. In this paperwork, the reference image and the real-time image captured from the camera is loaded in the image acquisition of LabVIEW. To process the acquired image, four different methods of kernels namely Arbitrary, Laplacian, Prewitt and Sobel methods are used to obtain an edge detection image. The edge detection images have stored and captured images are compared and the Root Mean Square Error is calculated to estimate the timing constraints to operate the traffic signal lights on a four-lane dynamically. LabVIEW graphical programming tools are used for the development of the scheme and simulation results are shown. Finally, the performance of the four methods analyzed using an image quality metric RMSE value to estimate the time in order to allow the vehicles in a particular direction and dynamically to switch them on and off control from one particular direction to another.