Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
The process of license plate recognition, including plate extraction based on edge statistics and morphology, character recognition, and filtering of connected components. The document also discusses the use of bilinear interpolation and template matching for character normalization and recognition. The results show a character recognition rate of 51.98% with some frequently confused characters.
Typology: Slides
1 / 20
This page cannot be seen from the preview
Don't miss anything!
In this phase License Plate attached to vehicle is extracted from the image based on Edge Statistics and Morphology.
License Plate Recognition Rate is 91% (100 Images).
Average processing time is 2 seconds.
docsity.com
Character Recognition:
The second phase is to recognize characters on the extracted license plate.
The characters on license plate are alpha numerals.
docsity.com
Components satisfying below conditions are accepted only.
Heigth_Object>=Heigth_of_Plate/3 &
Width_Object<Width_of_Plate/
docsity.com
Result of applying these constraint is that we have eliminated long lines and very small characters other than license plate number.
Along with the above irrelevant details some useful information has also been lost such as shape of characters.
To get rid of it objects are extracted from the unfiltered plate image.
docsity.com
Phase 2(Contd.)
docsity.com
Phase 2(Contd.)
To ease the process of identification
and achieve a high success rate it
is preferable to segment characters.
Individual characters on the plate are
extracted using 8-connected component
analysis.
docsity.com
Fig 5: Templates
docsity.com
Phase 2(Contd.)
docsity.com
Phase 2(Contd.)
Bilinear Interpolation:
In bilinear interpolation four nearest neighbor of a point are used for gray level assignment.
v (x', y') = ax' + by' + cx'y' + d
v (x', y') is the gray level assigned to a point (x', y') using bilinear interpolation
docsity.com
Phase 2(Contd.)
Template Matching:
For template matching the cross
correlation technique is used.
The cross-correlation (or sometimes "cross-covariance") is a measure of similarity of two signals.
Gives maximum value at the location where it finds maximum correspondance.
docsity.com
Phase 2(Contd.)
Character Recognition rate is 51.98 %.
278 characters were correctly recognized out of 535 characters.
Frequently confused characters are (B, D, and O) (Z, 7) (P, R) (A, 4)
docsity.com
Correlation was very sensitive to font size and frequently gave back wrong results
due to different size license plates.
In our case as images are not taken from fixed distances so we have to deal with scaling.
So this technique has been rejected due
to its inability to deal with scaling.
docsity.com
References: Bai Hongliang and Liu Changping ” A hybrid License Plate Extraction Method Based onEdge Statistics and Morphology” Proceedings of the 17th International Conference on Pattern Recognition , Volume: 2, ppt 831- 834, 2004. M J. Ahmed, M Sarfaz, A. Zidouri, and K G. AI- System” in Electronics, Circuits and Systems, Volume :2, ppt 898 –Khatib ” License Plate Recognition 901, ICECS 2003. Tran Duc Duan, Tran Le Hong Du, Tran Vinh Phuoc and Nguyen Viet Hoang “Building an Automatic Vehicle License- Plate Recognition System ” in Intl. Conf. in Computer Science, Can Tho, Vietnam, 2005. John C.M.Lee, W.K.Wong and H.S.Fong, “Automatic Character Recognition for Movingand Stationary Vehicles and Containers” in International Joint Conference on Neural Networks, Volume: 4, ppt 2824-2828 ,1999. Hans A. Hegt, Ron J. De la Haye and Nadeem A. Khan” A High Performance License PlateRecognition System” in IEEE International Conference on Systems, Man, and Cybernetics, 1998. Shyang-Lih Chang, Li-Shien Chen, Yun-Chung Chung, and Sei-Wan ChenLicense Plate Recognition ” in IEEE Transaction on Intelligent Transportation Systems, ,” Automatic Vol5, 2004. http://www.zemris.fer.hr/LicensePlates/english/results.shtml
docsity.com
docsity.com
