Title :
A registration method for vehicle license plate image
Author :
Li, Nan ; Lu, Xiaobo
Author_Institution :
Sch. of Transp., Southeast Univ., Nanjing, China
Abstract :
Image registration is the precondition of image enhance and image reconstruction which uses sequence low-resolution images. This paper presents a new intensity-based image registration, which combine feature based and statistical based methods. The proposed method consists of three steps. First, the center of gravity (CG) method is used to estimate the relative displacement. Then, the correlation method is used to estimate the rotation and zoom. Finally, the deformation function is established and the registration is processed. Simulation experimental results show the effectiveness of the proposed method and its superiority to the normal correlation methods in saving calculation.
Keywords :
feature extraction; image enhancement; image reconstruction; image registration; image resolution; statistical analysis; vehicles; center of gravity method; combine feature based method; correlation method; deformation function; image enhancement; image reconstruction; intensity-based image registration; sequence low-resolution images; statistical based method; vehicle license plate image; Accuracy; Biomedical imaging; Correlation; Image registration; Licenses; Pixel; Simulated annealing; center of gravity; correlation; feature point; registration; vehicle license plate images;
Conference_Titel :
Natural Computation (ICNC), 2010 Sixth International Conference on
Conference_Location :
Yantai, Shandong
Print_ISBN :
978-1-4244-5958-2
DOI :
10.1109/ICNC.2010.5584149