• No results found

Vol 7, No 6 (2017)

N/A
N/A
Protected

Academic year: 2020

Share "Vol 7, No 6 (2017)"

Copied!
7
0
0

Loading.... (view fulltext now)

Full text

(1)

Computer Science and Software Engineering

ISSN: 2277-128X (Volume-7, Issue-6)

2017

License Plate Localization Method Based on VerticalEdge

Detection

Neha Rana

MTech Scholar, Deenbandhu Chhotu Ram University of Science and Technology, Murthal, Sonepat, India

Pawan Kumar Dahiya

Assistant Professor, Deenbandhuchhotu Ram University of Science and Technology, Murthal, Sonepat, India DOI: 10.23956/ijarcsse/SV7I5/0337

Abstract: In our examination, a powerful algorithm for plate recognition which is able to recognize plates having different dimensions in unaccustomed and complicated backgrounds.License plate detection is first step in ANPR systems. The paper defines an algorithm to detect license plate.The plate is detected based on its edge density features. Integral edge image is calculated and then edge density conditions are verified.Two stage detection process is used and provides 97% accuracy.

Keywords:ANPR, CCA, SAICCA,

I. INTRODUCTION

ANPR comprises of three stages namely localization, segmentation followed by recognition of characters. There is no need of any hardware to be installed on vehicles, it works in real time mode, hence more economical.Based on government constraints on license plate dimension and color, algorithm can be developed.ANPR systems provides automated detection and recognizing of a vehicle. Thus, these systems when combined with a database, are used in various security applications’ systems takes image as an input and without human intervention plate number is detected and saved for future action.

In license plate localization, plate is located from the image captured by camera.The difficulties are vehicle speed, lightning conditions, heavy traffic.In this scientific era, security is major concern. Shape size and color are different for different countries [12,9,11,10]. Figure 1 illustrates example of Indian license plate. Generally, license plates are distinguished by high contrast between characters and background. Advancement in technology like infrared imaging and high-resolution cameras, and use of high reflective backgrounds in license plate manufacturing have improved the accuracy of ANPR systems.Hardware used adds to the accuracy of the system.

The contribution of the paper is as under:

 Algorithm to detect a license plate

 Analysis of performance of the technique

Fig.1: Sample license plate images[13]

A typical LPR system is composed of several hardware and software components as illustrated in Figure 2

(2)

ISSN: 2277-128X (Volume-7, Issue-6)

The rest of the paper is organized as under: Section 2: Literature Survey

Section 3: Proposed methodology Section 4: Software Used and Result Section 5:Conclusions

II. LITERATURE SURVEY

Lorita Angeline[3] inthe paper presents Signature Analysis and the improved conventional Connected Component Analysis (CCA) to design an automatic license plate localization and identification. A procedure called Euclidean Distance Transform is added to the conventional CCA in order to tackle the multiple bounding boxes that occurred. The developed algorithm, SAICCA achieved 92% successful rate, with 8% failed localization rate due to the restrictions such as insufficient light level, clarity and license plate perceptual information. Pushpendra Singh Chouhan[2] inthe paper presents a characteristic based methodology for localization of Indian number plates. Alphanumeric characters have specific characteristics in binary image. An approach for localization of license plate using those characteristics is presented. The results demonstrate an accuracy of 90%.BalazsEnyedi[8] in the paper proposed algorithms based on fast and classical image processing methods such as filtering, edge finding and adaptive thresholding, which do not incorporate any learning procedure, shall be discussed. Mohammad H. Dashtban[5] in the paper proposed a method that uses Hough Transform, morphological operator and skeletonizing to provide and input to neural network. Segment by segment the input streams into an intelligent control unit. This method achieves 91% accuracy. SatadalSaha[7] proposed technique which is based on color based segmentation of vehicle images and identification of potential license plate regions. True license plates are finally localized based on four spatial and horizontal contrast features. The technique successfully localizes the actual license plates in 73.4% images. Arya P. Unnikrishnan[1] in the paper presents a concept which is the combination of image processing and genetic algorithms. Here is a proposal for a design of a new genetic algorithm, which is introduced to detect the license plate location. It is done by converting the taken image into binary format from gray-scale image. Also connected component analysis is implemented to detect the candidate objects in that given image. XiaojunZhai[6] proposed method which is based on morphological open and close operations where different Structuring Elements (SE) are used to maximally eliminate non-plate region and enhance plate region.

III. METHODOLOGY

In this research work, the method proposed to localize license plate, consist of 3 main parts. A.Binarization of the image

B.Vertical edge detection C.Sliding window technique

An input image submitted to the system is first examined and processed to obtain the vehicle license plate region.

3.1 BINARIZATION

To extract the license plate first the RGB image is binarized and converted to binary.After that adaptive thresholding is used to enhance the contrast details in the image.

Fig.3: Flow diagram of proposed method plate localization

This algorithm is based on the contrast between the gray scale values[4]. They have used adaptive thresholding for binarization, an unwanted line elimination algorithm followed by vertical edge detection using 2 * 4 mask and finally candidate region extraction for license plate detection [4].

3.2 ADAPTIVE THRESHOLDING

Thresholding cannot provide us better results under various illuminant conditions. Hence adaptive thresholding is used in which each pixel value is compared with the threshold value calculated using its neighboringpixels. Adaptive thresholding results in 1 when the pixel value is greater than threshold value and vice versa.The

(3)

ISSN: 2277-128X (Volume-7, Issue-6)

rectangular window size can be calculated from the width of the input image. It can be one eighths of the image width. [4]

s=N/8

Where N is the width of the image and s is the local windowrectangle size s Xs.

So, for any center pixel location (i, j), the rectangle corner coordinates are s1, j+s1), (i+s1, j+s1), (i+s1, j-s1) and

(i-s1, j-s1) Thresholdt(i, j) for each pixel can be computed by using the following equation [13].

t (x,y)=(1—T)∗windowsum

Where T is a constant (T = 0.15) which is the best value for the all types of images after testing on many images [4].

3.3 VERTICAL EDGE DETECTION

Afterbinarization,nextstepistocalculateverticaledgeswhichcanbe computed by using spatial filtering between a predefined mask and the binary image. A proposed mask of size 2 X 4 can be defined as followswhich

1 -2 0 1

1 -2 0 1

Fig.4: A 2 X 4 mask for vertical edge detection[4]

3.4 SLIDING WINDOW TECHNIQUE

A sliding window is used to mark all the windows that satisfies the edge density conditions. Edge density is evaluated on integral image calculated from the input image. After that window size is increased and again condition is checked for the windows that were marked in earlier stage. Sliding window size is increased in every stage until it reaches the size on input image. The output of this sliding window is our localized license plate. Sliding window detection has one fundamental weakness: it is very computational expensive to search and evaluate all possible window locations. To deal with this issue the number of evaluations is reduced using the translation and size steps which define how fast the window is moved and resized. These steps need to be chosen carefully because they represent a tradeoff between the speed and the localization robustness of the LPL method [4].

IV. SOFTWARE USED AND RESULT

Software: MATLAB Version R2015a: It is powerful software that provides an environment for numerical computation as well as graphical display of outputs. It is high-performance language for technical computing integrates computation, visualization, and programming in a simple way where problems and solutions are expressed in familiar mathematical notation.

(4)

ISSN: 2277-128X (Volume-7, Issue-6)

Fig.6: Gray image and column direction pixel gray value

Fig.7: Sobel edge detection technique

(5)

ISSN: 2277-128X (Volume-7, Issue-6)

Fig.9: Gray scale linear transformation

Fig.10: Filtered Image

(6)

ISSN: 2277-128X (Volume-7, Issue-6)

Fig.12: Plate binary value

Fig.13: License plate detected

Fig.14: First 3 images from base paper and last one is our proposed [4]

V. CONCLUSIONS

A robust method for license plate detection is presented in the paper. It exploits the fact that the license plate area contains rich edge and texture information. First, the vertical edges are extracted and the edge map is adaptively binarized. Then, the license plate candidates are detected using the two-stage detection process which consists from coarse and fine window detection. Finally, candidates are verified using the geometrical and textural properties. To improve the efficiency of the algorithm the concept of integral image was adopted.Comparison of various techniques is done below along with this technique represented in this paper [14].

Technique Approach Accuracy achieved Using Signature

Analysis

Signature analysis, CCA, Euclidean distance transform

92%

Using alphanumeric characters

Pixel density evaluation, Filtering

90%

Novel approach Pixel connectivity, Hough transform, Morphological operator, Skeletonizing

91%

Using Genetic algorithm

CCA, Genetic crossover operators, Color feature

(7)

ISSN: 2277-128X (Volume-7, Issue-6)

Using morphological operators

Morphological operation with different structuring elements

91%

Using vertical edge detection

Vertical edge detection based on sliding window technique

97%

Fig.15: Table of performance comparison of various techniques [14]

REFERENCES

[1] Arya P. Unnikrishnana, Roshini Romeo, Fabeela Ali Rawther , ―License Plate Localization using Genetic Algorithm including Color Feature Extraction,‖ in International Conference on Emerging Trends in Engineering, Science and Technology (ICETEST),2015.

[2] Pushpendra Singh Chouhan , V. K. Govindan, ―Localization of License Plate using Characteristics of Alphanumeric Characters,‖ (IJCSIT) International Journal of Computer Science and Information Technologies, Vol. 5 (3) , 3407-3409, 2014.

[3] Lorita Angeline, Hou Pin Yoong, Hui Keng Lau, Ismail Saad and Kenneth Tze Kin Teo, ―Automatic License Plate Localisation And Identification Via Signature Analysis,‖ ICTACT Journal on Image and Video Processing, Volume: 04, Issue: 03, 2014.

[4] Peter Tarabek,‖ A Real-Time License Plate Localization Method Based on Vertical Edge Analysis,‖ in Proceedings of the Federated Conference on Computer Science and Information Systems pp. 149–154,2012. [5] Mohammad H. Dashtban, ZehraDashtban, Hassan Bevrani,‖ A novel approach for license plate localization and

recognition‖, in International journal of computer applications (0975-8887),volume 26, No. 11, 2011.

[6] XiaojunZhai, FaycalBenssali and SoodamaniRamalingam, ―License Plate Localisation based on Morphological Operations,‖ in 11th Int. Conf. Control, Automation, Robotics and Vision Singapore, 2010.

[7] SatadalSaha , Subhadip Basu , MitaNasipuri , Dipak Kumar Basu,‖ An Offline Technique for Localization of License Plates for Indian Commercial Vehicles ,‖ in National Conference on Computing and Communication Systems (COCOSYS-09) CS10,2009.

[8] Bal´azsEnyedi, Lajos Konyha, K´alm´anFazeka, ―Real Time Number Plate Localization Algorithms,‖ Journal of ELECTRICAL ENGINEERING, VOL. 57, NO. 2, 69–77, 2006.

[9] T. Naito, T. Tsukada, K. Yamada, K. Kozuka, and S. Yamamoto, ―Robust License-Plate Recognition Method for Passing Vehicles Under Outside Environment. Vehicular Technology,‖ IEEE Transactions On 49(6):2309– 2319,2000.

[10] T. Sirithinaphong And K. Chamnongthai,‖ The Recognition Of Car License Plate For Automatic Parking System‖ in Signal Processing And Its Applications,ISSAP , proceedings Of The Fifth International Symposium, Volume 1, Pages 455–457,1999.

[11] J. A G Nijhuis, M.H. Ter Brugge, K. A. Helmholt, J. P W Pluim, Spaanenburg, R. S. Venema, And M.A. Westenberg,‖ Car License Plate Recognition with Neural Networks and Fuzzy Logic,‖ in Neural Network 1995 Proceedings, IEEE International Conference, Volume 5, Pages 2232–2236,1995.

[12] EunRyung Lee, PyeoungKee Kim, And Hang Joon Kim, ―Automatic Recognition ofA Car License Plate Using Color Image Processing,‖ in ICIP, IEEE International Conference, Volume 2, Pages 301–305,1994.

[13] Google Image accessed on 28.5.17.

References

Related documents

Although theoretically the likelihood of finding evidence that dumped imports have in- jured the domestic industry should fall as the industry increases its output, the results from

Most algorithms for large item sets are related to the Apri- ori algorithm that will be discussed in Chapter IV-A2. All algorithms and methods are usually based on the same

The UNO Assessment Committee is responsible for guiding the process of campus-wide academic assessment of student learning, and to that end it conducts regular reviews of

3This result is explained by the fact that under Bertrand competition foreign investment in R&D has a negative (indirect) strategic effect on foreign firms'

[r]

Although total labor earnings increase with the unskilled unions’ bargaining power, we can say nothing when the increase in production is due to stronger skilled unions, since

Using a nationwide database of hospital admissions, we established that diverticulitis patients admitted to hospitals that encounter a low volume of diverticulitis cases have

According to CoreLogic, Inc., during the 12 months ending September 2012, sales of new single- family homes, townhomes, and condo - miniums totaled 5,075, down nearly 3