Morphology Based Approach for Number Plate Extraction
Number Plate Recognition identifies vehicle number without human intervention. It is a computer vision application and it has many important applications. The proposed system consists of two parts: number plate area extraction and character identification
- PDF / 281,946 Bytes
- 8 Pages / 439.37 x 666.142 pts Page_size
- 34 Downloads / 251 Views
Abstract Number Plate Recognition identifies vehicle number without human intervention. It is a computer vision application and it has many important applications. The proposed system consists of two parts: number plate area extraction and character identification. In this paper, morphological operation-based approach is presented for number plate area extraction. Effective segmentation of characters is done after plate area extraction. Histogram-based character segmentation is a simple and efficient technique used for segmentation. Template matching approach is used for character extraction. Number plate with variable character length poses limitation on number identification in earlier reported literature. This is taken care of using histogram-based character segmentation method.
⋅
Keywords Histogram Morphological operations Template matching Thresholding
⋅
⋅
Number plate extraction
⋅
1 Introduction Every country has specific vehicle identification system. These systems are used in the traffic control and surveillance systems, security systems, toll collection at toll plaza and parking assistance system, etc. Human eye can easily recognize these number plates, but designing automated system for this task has many challenges. Blur, unequal illumination, background and foreground color and also many natural phenomena like rain fall, dust in air may create problem in number extraction. Also number plate standards are different for each country, therefore large number of variations are obtained in parameters like, location of number plate, area of number plate and characters, font and size used for numbers and characters (standard font is Chetan Pardeshi (✉) ⋅ Priti Rege Department of Electronics and Telecommunication, College of Engineering Pune, Pune, India e-mail: [email protected] Priti Rege e-mail: [email protected] © Springer Science+Business Media Singapore 2017 S.C. Satapathy et al. (eds.), Proceedings of the International Conference on Data Engineering and Communication Technology, Advances in Intelligent Systems and Computing 469, DOI 10.1007/978-981-10-1678-3_2
11
12
Chetan Pardeshi and Priti Rege
Arial Black), background color (white, yellow or black) and foreground color (black or red), etc., which make the task of number plate extraction difficult. Number of applications of license plate identification can be listed as parking assistance facility during ticket collection, unmanned toll collection at toll booths, traffic surveillance system, tracking vehicles during signal violation, vehicle’s marketing research. Aim of this paper is to implement an efficient method for number plate extraction. Algorithm proposed in this paper identifies characters present on single line number plate with variable character length. Rege and Chandrakar [1] has separated text image in document images using run length searing algorithm and boundary perimeter detection. Morphological operation and bounding box analysis are used by Patel et al. [2]. Owamoyo et al. [3] used Sobel filter along with morphological operat
Data Loading...
