一、外文资料
Licen Plate Recognition Bad On Prior Knowledge
Abstract - In this paper, a new algorithm bad on improved BP (back propagation) neural network for Chine vehicle licen plate recognition (LPR) is described. The propod approach provides a solution for the vehicle licen plates (VLP) which were degraded verely. What it remarkably differs from the traditional methods is the application of prior knowledge of licen plate to the procedure of location, gmentation and recognition. Color collocation is ud to locate the licen plate in the image. Dimensions of each character are constant, which is ud to gment the character of VLPs. The Layout of the Chine VLP is an important feature, which is ud to construct a classifier for recognizing. The experimental results show that the improved algorithm is effective under the condition that the licen plates were degraded verely.
Index Terms - Licen plate recognition, prior knowledge, vehicle licen plates, neural network.
I. INTRODUCTION
Vehicle Licen-Plate (VLP) recognition is a very interesting but difficult problem. It is important in a nu
mber of applications such as weight-and-speed-limit, red traffic infringement, road surveys and park curity [1]. VLP recognition system consists of the plate location, the characters gmentation, and the characters recognition. The tasks become more sophisticated when dealing with plate images taken in various inclined angles or under various lighting, weather condition and cleanliness of the plate. Becau this problem is usually ud 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 increa the recognition rate by using some specific features of local VLPs and establishing some constrains on the position, distance from the camera to vehicles, and the inclined angles. In addition, neural network was ud to increa the recognition rate [6], [7] but the traditional recognition methods ldom consider the prior knowledge of the local VLPs. In this paper, we propod a new improved learning method of BP algorithm bad on specific features of Chine VLPs. The propod algorithm overcomes the low speed convergence of BP neural network [8] and remarkable increas the recognition rate especially under the condition that the licen plate images were degrade verely.
II. SPECIFIC FEATURES OF CHINESE VLPS
A. Dimensions
According to the guideline for vehicle inspection [9], all licen 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 to vehicles and the inclined angles are constant, so all characters of the licen 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. Tho features can be ud to locate the plate and gment the individual character. B. Color collocation of the plate
There are four kinds of color collocation for the Chine vehicle licen plate .The color collocations are shown in table I.
TABLE I
山西省分数线
Moreover, military vehicle and police wagon plates contain a red character which belongs to a specific character t. This feature can be ud to improve the recognition rate.
C. Layout of the Chine VLPS
The criterion of the vehicle licen plate defines the characters layout of Chine licen plate. All standard licen plates contain Chine characters, numbers and letters which are shown in Fig.1. The first one is a Chine character which is an abbreviation of Chine provinces. The cond 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 venth ones are numbers ranging from 0 to 9 only. However the first or the venth ones may be red characters in special plates (as shown in Fig.1). After gmentation process the individual character is extracted. Taking advantage of the layout and color collocation prior knowledge, the individual character will enter one of the class: abbreviations of Chine provinces t, letters t, letters or numbers t, number t, special characters t.
(a)Typical layout
(b) Special character
Fig.1 The layout of the Chine licen plate
III. THE PROPOSED ALGORITHM
This algorithm consists of four modules: VLP location, character gmentation, character classification and character recognition. The main steps of the flowchart
四月初一of LPR system are shown in Fig. 2.
Firstly the licen plate is located in an input image and characters are gmented. Then every individual character image enters the classifier to decide which class it belongs to, and finally the BP network decides which character the character image reprents.
fullfillFig.2 The flowchart of LPR system
南极的英文A. Preprocessing the licen plate
1) VLP Location
This process sufficiently utilizes the color feature such as color collocation, color centers and distrib
ution in the plate region, which are described in ction II. The color features can be ud 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 plate are extracted. The process is described as followed:
农村贫困户申请书(1)
(2)
Here, the Gaussian variance is t 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 * M (T<0.5). Median filter is ud to prerve the edge gradient and eliminate isolated noi of the binary image. An N * N rectangle median filter is t, and N reprents the odd integer mostly clo to W.
Morphology closing operation can be ud to extract the candidate region. The confidence degree o
f candidate region for being a licen plate is verified according to the aspect ratio and areas. Here, the aspect ratio is t between 1.5 and 4 for the reason of inclination. The prior knowledge of color collocation is ud to locate plate region exactly. The locating process of the licen plate is shown in Fig. 4. Characters edge detection职场关系
Binary image gmenting
可行性报告怎么写Candidate image detection
Vehicle plate extraction
Fig. 4 The whole process of locating licen plate
2) Character gmentation
This part prents an algorithm for character gmentation bad on prior knowledge, using character width, fixed number of characters, the ratio of height to width of a character, and so on. The flowchart of the character gmentation is shown in Fig. 5.
Licen plate image
preprocessing
Obtain binary image
Vertical projection
Eliminate space mark
中胚型Fig. 5 The flowchart of the character gmentation
Firstly, preprocess the licen the plate image, such as uneven illumination correction, contrast enhancement, incline correction and edge enhancement operations; condly, eliminating space mark which appears between the cond character and the third character; thirdly, merging the gmented fragments of the characters. In China, all standard licen plates contain only 7 characte
rs (e Fig. 1). If the number of gmented characters is larger than ven, the merging process must be performed. Table II shows the merging process. Finally, extracting the individual character’image bad on the number and the width of the character. Fig. 6 shows the gmentation results. (a) The incline and broken plate image, (b) the incline and distort plate image, (c)the rious fade plate image, (d) the smut licen plate image.
