p.2515
p.2519
p.2524
p.2530
p.2534
p.2538
p.2542
p.2549
p.2553
Quadrilateral Extraction Method and its Application on Parking Cell Detection
Abstract:
This paper presents a quadrilate extraction method based on four vertices of the image object to solve the problrm of object area inaccuracy depending on conventional rectangular segmentation method. Firstly, four vertices are given by the mouse-click depending on quadilateral object spatial feature; and then the quadrilateral object shape feature is estimated by each side of the quadriangle computing by straight-line formulation of two adjacent vertices of object in order to acquare the quadrialeratal object inernal pixel. At last, the pixels of quadrialeratel external region are cleaned. Arbitrary quadrilateral region segmentation is achieved accurately. The method can be used for any quadrilateral region segmentation based on image processing, The experiments on actual color parking images show any parking cell template extraction accuracy reached 100% and parking cell state detection accuracy is 99.8%. The result verifies the algorithm effectiveness and robustnes
Info:
Periodical:
Pages:
2534-2537
Citation:
Online since:
December 2012
Authors:
Price:
Сopyright:
© 2013 Trans Tech Publications Ltd. All Rights Reserved
Share:
Citation: