Character Extraction of Vehicle Plate Directly from Compressed Video Stream

Character extraction of vehicle plate is conducted directly in MPEG compressed domain. Firstly, vehicle plate region is detected by thresholding the weighted horizontal texture energy, in which adaptive threshold is used, followed by a series of post processing, e. g. morphological operations, connected component analysis, filtering with aspect ratio constrains. Secondly, block’s DC+2AC texture intensity projection profile is used to implement the character of vehicle plate localization accurately. Then after minimal decompression processing and contrast adjustment on the located region, OTSU thresholding technique is used for character extraction. Experiment results show the effectiveness of the proposed algorithm.

[1]  Anil K. Jain,et al.  Automatic Caption Localization in Compressed Video , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  Yang Gaobo,et al.  A Fast and Accurate Moving Object Extraction Scheme in the MPEG Compressed Domain , 2007, Fourth International Conference on Image and Graphics (ICIG 2007).

[3]  Seong-Whan Lee,et al.  Text extraction in MPEG compressed video for content-based indexing , 2000, Proceedings 15th International Conference on Pattern Recognition. ICPR-2000.

[4]  Anil K. Jain,et al.  Text information extraction in images and video: a survey , 2004, Pattern Recognit..

[5]  Julinda Gllavata,et al.  Extracting textual information from images and videos for automatic content-based annotation and retrieval , 2007 .

[6]  Qian Huang,et al.  Character extraction of license plates from video , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[7]  N. Otsu A threshold selection method from gray level histograms , 1979 .