1、 一、外 文 资 料 License Plate Recognition Based On Prior Knowledge Abstract - In this paper, a new algorithm based on improved BP (back propagation) neural network for Chinese vehicle license plate recognition (LPR) is described. The proposed approach provides a solution for the vehicle license plates (V
2、LP) which were degraded severely. What it remarkably differs from the traditional methods is the application of prior knowledge of license plate to the procedure of location, segmentation and recognition. Color collocation is used to locate the license plate in the image. Dimensions of each characte
3、r are constant, which is used to segment the character of VLPs. The Layout of the Chinese VLP is an important feature, which is used to construct a classifier for recognizing. The experimental results show that the improved algorithm is effective under the condition that the license plates were degr
4、aded severely. Index Terms - License plate recognition, prior knowledge, vehicle license plates, neural network. I. INTRODUCTION Vehicle License-Plate (VLP) recognition is a very interesting but difficult problem. It is important in a number of applications such as weight-and-speed-limit, red traffi
5、c infringement, road surveys and park security 1. VLP recognition system consists of the plate location, the characters segmentation, and the characters recognition. These tasks become more sophisticated when dealing with plate images taken in various inclined angles or under various lighting, weath
6、er condition and cleanliness of the plate. Because this problem is usually used in real-time systems, it requires not only accuracy but also fast processing. Most existing VLP recognition methods 2, 3, 4, 5 reduce the complexity and increase the recognition rate by using some specific features of lo
7、cal VLPs and establishing some constrains on the position, distance from the camera to vehicles, and the inclined angles. In addition, neural network was used to increase the recognition rate 6, 7 but the traditional recognition methods seldom consider the prior knowledge of the local VLPs. In this
8、paper, we proposed a new improved learning method of BP algorithm based on specific features of Chinese VLPs. The proposed algorithm overcomes the low speed convergence of BP neural network 8 and remarkable increases the recognition rate especially under the condition that the license plate images w
9、ere degrade severely. II. SPECIFIC FEATURES OF CHINESE VLPS A. Dimensions According to the guideline for vehicle inspection 9, all license plates must be rectangular and have the dimensions and have all 7 characters written in a single line. Under practical environments, the distance from the camera
10、 to vehicles and the inclined angles are constant, so all characters of the license plate have a fixed width, and the distance between the medium axes of two adjoining characters is fixed and the ratio between width and height is nearly constant. Those features can be used to locate the plate and se
11、gment the individual character. B. Color collocation of the plate There are four kinds of color collocation for the Chinese vehicle license plate .These color collocations are shown in table I. TABLE I Category of license plate Color collocation small horse power plate blue background and white char
12、acters motor truck plate yellow background and black characters military vehicle and police wagon plate black background and the white characters embassy vehicle plate white background and black characters Moreover, military vehicle and police wagon plates contain a red character which belongs to a
13、specific character set. This feature can be used to improve the recognition rate. C. Layout of the Chinese VLPS The criterion of the vehicle license plate defines the characters layout of Chinese license plate. All standard license plates contain Chinese characters, numbers and letters which are sho
14、wn in Fig.1. The first one is a Chinese character which is an abbreviation of Chinese provinces. The second one is a letter ranging from A to Z except the letter I. The third and fourth ones are letters or numbers. The fifth to seventh ones are numbers ranging from 0 to 9 only. However the first or
15、the seventh ones may be red characters in special plates (as shown in Fig.1). After segmentation process the individual character is extracted. Taking advantage of the layout and color collocation prior knowledge, the individual character will enter one of the classes: abbreviations of Chinese provi
16、nces set, letters set, letters or numbers set, number set, special characters set. (a)Typical layout (b) Special character Fig.1 The layout of the Chinese license plate III. THE PROPOSED ALGORITHM This algorithm consists of four modules: VLP location, character segmentation, character classification
17、 and character recognition. The main steps of the flowchart of LPR system are shown in Fig. 2. Firstly the license plate is located in an input image and characters are segmented. Then every individual character image enters the classifier to decide which class it belongs to, and finally the BP netw
18、ork decides which character the character image represents. Fig.2 The flowchart of LPR system Image acquisition Plate location Characters segmentation segmentation classifier Chinese character Letter Letter or number Number Special character Characters recognition 辽 BB092 警 Chinese character charact
19、er Letter Letter or number Special red character 辽 BA9083 Chinese character character Letter Letter or number Number A. Preprocessing the license plate 1) VLP Location This process sufficiently utilizes the color feature such as color collocation, color centers and distribution in the plate region,
20、which are described in section II. These color features can be used to eliminate the disturbance of the fake plate s regions. The flowchart of the plate location is shown in Fig. 3. Fig.3 The flowchart of the plate location algorithm The regions which structure and texture similar to the vehicle pla
21、te are extracted. The process is described as followed: (1) (2) Here, the Gaussian variance is set to be less than W/3 (W is the character stroke width), so 1P gets its maximum value M at the center of the stroke. After convolution, binarization is performed according to a threshold which equals T *
22、 M (T MaxF For each character segments Calculate the medium point iM For each two consecutive medium points Calculate the distance KD Calculate the minimum distance Merge the character segment k and the character segment k +1 NF = NF - 1 End of algorithm where Nf is the number of character segments,
23、 MaxF is the number of the license plate, and i is the index of each character segment. The medium point of each segmented character is determined by: （ 3） where 1iS is the initial coordinates for the character segment, and 2iS is the final coordinate for the character segment. The distance between
24、two consecutive medium points is calculated by: （ 4） Fig.6 The segmentation results B. Using specific prior knowledge for recognition The layout of the Chinese VLP is an important feature (as described in the section II), which can be used to construct a classifier for recognizing. The recognizing p
25、rocedure adopted conjugate gradient descent fast learning method, which is an improved learning method of BP neural network10. Conjugate gradient descent, which employs a series of line searches in weight or parameter space. One picks the first descent direction and moves along that direction until
26、the minimum in error is reached. The second descent direction is then computed: this direction the “ conjugate direction” is the one along which the gradient does not change its direction will not “ spoil” the contribution from the previous descent iterations. This algorithm adopted topology 625-35-
27、N as shown in Fig. 7. The size of input value is 625 (25*25 ) and initial weights are with random values, desired output values have the same feature with the input values. Fig. 7 The network topology As Fig. 7 shows, there is a three-layer network which contains working signal feed forward operatio
28、n and reverse propagation of error processes. The target parameter is t and the length of network output vectors is n. Sigmoid is the nonlinear transfer function, weights are initialized with random values, and changed in a direction that will reduce the errors. The algorithm was trained with 1000 i
29、mages of different background and illumination most of which were degrade severely. After preprocessing process, the individual characters are stored. All characters used for training and testing have the same size (25*25 ).The integrated process for license plate recognition consists of the followi
30、ng steps: 1) Feature extracting The feature vectors from separated character images have direct effects on the recognition rate. Many methods can be used to extract feature of the image samples, e.g. statistics of data at vertical direction, edge and shape, framework and all pixels values. Based on
31、extensive experiments, all pixels values method is used to construct feature vectors. Each character was reshaped into a column of 625 rows feature vector. These feature vectors are divided into two categories which can be used for training process and testing process. 2) Training model The layout o
32、f the Chinese VLP is an important feature, which can be used to construct a classifier for training, so five categories are divided. The training process of numbers is shown in Fig. 8. Fig. 8 The architecture of a neural network for character recognition As Fig. 8 shows, firstly the classifier decid
33、es the class of the input feature vector, and then the feature vector enters the neural network correspondingly. After the training process the optimum parameters of the net are stored for recognition. The training and testing process is summarized in Fig. 9. (a) Training process (b)Testing process
34、Fig.9 The recognition process Input character vector for recognition 特征向量 Neural network output Input character vector for recognition 特征向量 Neural network Target output error 3) Recognizing model After training process there are five nets which were completely trained and the optimum parameters were
35、 stored. The untrained feature vectors are used to test the net, the performance of the recognition system is shown in Table III. The license plate recognition system is characterized by the recognition rate which is defined by equation (5). Recognition rate =(number of correctly read characters)/ (
36、number of found characters) (5) TABLE III Class Recognition Number 99.5% Letter 97.4% Chinese character 96% Number and letter 97.3% Special character 98.2% IV. COMPARISON OF THE RECOGNITION RATE WITH OTHER METHODS In order to evaluate the proposed algorithm, two groups of experiments were conducted.
37、 One group is to compare the proposed method with the BP based recognition method 11. The result is shown in table IV. The other group is to compare the proposed method with the method based on SVM 12.The result is shown in table V. The same training and test data set are used. The comparison result
38、s show that the proposed method performs better than the BP neural network and SVM counterpart. TABLE IV Method Chinese character Number Letter Our method 96% 99.5% 97.4% BP 94 5% 97.6% 89.8% TABLE V Method Chinese character Number Letter Our method 96% 99.5% 97.4% SVM 93 7% 99.5% 95.7% V. CONCLUSIO
39、N In this paper, we adopt a new improved learning method of BP algorithm based on specific features of Chinese VLPs. Color collocation and dimension are used in the preprocessing procedure, which makes location and segmentation more accurate. The Layout of the Chinese VLP is an important feature, wh
40、ich is used to construct a classifier for recognizing and makes the system performs well on scratch and inclined plate images. Experimental results show that the proposed method reduces the error rate and consumes less time. However, it still has a few errors when dealing with specially bad quality
41、plates and characters similar to others. This often takes place among these characters (especially letter and number):3 8 4 A 8 B D 0. In order to improve the incorrect recognizing problem we try to add template-based model 13 at the end of the neural network. REFERENCES 1 P. Davies, N. Emmottand N.
42、 Ayland,“ License Plate Recognition Technology for Toll Violation Enforcement” Proceedings of IEE Colloquium on Image analysis for Transport Applications, Vol. 035, pp.7/1-7/5, February 16, 1990. 2 V. Koval,V. Turchenko,V. Kochan, A. Sachenko, G. Markowsky,“ Smart. License Plate Recognition System B
43、ased on Image Processing Using Neural Network” IEEE International Workshop on Intelligent Data Acquisition and Advanced Computing System: Technology Applications 8-10 September 2003. 3 Abdullah, S.N.H.S.; Khalid, M.； Yusof, R.; Omar, K.“ License Plate Recognition using multi-cluster and Multilayer N
44、eural Networks” Information and Communication Technologies, 2006. ICTTA 06. 2nd Volume 1, 24-28 April 2006 Page(s):1818 1823. 4 Nathan, V.S.L.; Ramkumar, J.; Kamakshi Priya, S.“ New approaches for license plate recognition system” Intelligent Sensing and Information Processing, 2004. Proceedings of
45、International Conference on 2004 Page(s):149 152. 5 Mei Yu; Yong Deak Kim,“ An approach to Korean license plate recognition based on vertical edge matching” Volume 4, 8-11 Oct. 2000 Page(s):2975 - 2980 vol.4. 6 Tindall, D.W.” Application of neural network techniques to automatic licence plate recogn
46、ition” Security and Detection, 1995., European Convention on 16-18 May 1995 Page(s):81 85. 7 Aghdasi, F.; Ndungo, H. “ Automatic licence plate recognition system” AFRICON, 2004. 7th AFRICON Conference in AfricaVolume 1, 2004 Page(s):45 - 50 Vol.1 8 Richard O.Duda Peter E.Hart David G.Stork,“ Pattern Classification Second Edition” PP 333
