Open AccessA New Technique Based on Intensity Distribution Graphs for the Extraction of Traffic Flow Rate and Vehicle Speed
Author(s) -
G. Abo Samra and O. AbdulKader G. Abo Samra and O. AbdulKader
Publication year - 2012
Language(s) - English
Resource type - Journals
ISSN - 1658-6336
DOI - 10.4197/comp.1-2.1
Subject(s) - hough transform , computer vision , computer science , artificial intelligence , grayscale , background subtraction , frame rate , line (geometry) , orientation (vector space) , region of interest , optical flow , pixel , real time computing , simulation , image (mathematics) , mathematics , geometry
This paper introduces a new visual system which allowsanalysis of traffic behavior on freeways and highways, using a videocamera mounted on a relatively high place (such as a bridge). A newtechnique based on spatial and temporal intensity distribution isintroduced to determine the road traffic parameters: average speed(AVSP) and average flow rate (AVFL). After performing many stepsincluding automatic detection of the Region Of Interest (ROI) andcamera orientation, background subtraction and Inverse PerspectiveMapping (IPM), an Intensity Distribution Graph (IDG) is built for thecurrent frame. The (IDG) is then transformed into a single GrayscaleImage Line (GIL). A certain number (N) of GILs representing Nframes are concatenated to build a Vehicle Tracking Diagram image(VTD) which gives a whole view of the traffic flow in the ROI duringa specific period T. Using the VTD we can visualize the vehiclestrajectories, speeds and estimate AVFL. By applying edge detectionand Hough Line transform on the VTD, AVSP is estimated. Thesystem has been implemented using MATLAB as an offlineapplication. The results were encouraging, reaching 97% accuracy forthe AVSP and AVFL. Further code optimization is needed for onlinemonitoring.