CN108537223B - A license plate detection method, system and device and storage medium - Google Patents

Abstract

The invention discloses a license plate detection method, system and device, and a computer-readable storage medium. The method includes: acquiring a test sample, and extracting edge features, HOG features and HSV features of the test sample; calculating according to the edge features The Euclidean distance between the test sample and each character in the training set, and the character corresponding to the minimum value of all the Euclidean distances is determined as the target character; the license plate type is obtained through the training set, and the multi-class logistic regression algorithm and the HOG are used. The feature determines the target license plate type; obtains the license plate color feature through the training set, constructs an SVM classifier according to the license plate color feature, and inputs the HSV feature into the SVM classifier to obtain the target license plate color feature; The type of the target license plate and the color feature of the target license plate determine the license plate detection result of the test sample. The license plate detection method provided by the invention improves the accuracy of the license plate detection.

Description

A license plate detection method, system and device and storage medium

technical field

The present invention relates to the technical field of image processing, and more particularly, to a license plate detection method, system and device, and a computer-readable storage medium.

Background technique

In recent years, license plate recognition equipment has been widely used in parking lots, urban roads and other areas to automatically capture and recognize vehicle license plates. The license plate detection method in the prior art mainly detects a single feature, and the detection accuracy is often low in complex environments such as too strong noise, blur, half occlusion, and similar license plates.

Therefore, how to improve the license plate detection accuracy is a problem to be solved by those skilled in the art.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a license plate detection method, system and device and a computer-readable storage medium, which improve the license plate detection accuracy.

To achieve the above purpose, the present invention provides a license plate detection method, comprising:

Obtain a test sample, and extract the edge feature, HOG feature and HSV feature of the test sample;

Calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and determine the character corresponding to the minimum value of all the Euclidean distances as the target character;

Obtain the license plate type through the training set, and use the multi-class logistic regression algorithm and the HOG feature to determine the target license plate type;

Obtain the license plate color feature through the training set, construct an SVM classifier according to the license plate color feature, and input the HSV feature into the SVM classifier to obtain the target license plate color feature;

The license plate detection result of the test sample is determined according to the target character, the type of the target license plate, and the color feature of the target license plate.

Wherein, calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and determine the character corresponding to the minimum value of all the Euclidean distances as the target character, including:

Calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and normalize all the Euclidean distances;

The first basic probability distribution of the Euclidean distance after normalization is calculated, and the character corresponding to the maximum value among all the basic probabilities is determined as the target character.

Wherein, the multi-class logistic regression algorithm and the HOG feature are used to determine the target license plate type, including:

Using the multi-class logistic regression algorithm according to the HOG feature, the second basic distribution probability of the license plate type is calculated, and the license plate type corresponding to the maximum value among all the basic probabilities is determined as the target license plate type.

Wherein, the HSV feature is input into the SVM classifier to obtain the target license plate color type, including:

Using the SVM classifier according to the HSV feature, the third basic probability distribution of the license plate color type is calculated, and the license plate color type corresponding to the maximum value among all the basic probabilities is determined as the target license plate color type.

Among them, it also includes:

The trust degree of the license plate detection result is calculated according to the first basic probability distribution, the second basic probability distribution and the third basic probability distribution according to a preset fusion rule.

Wherein, the distribution of the first basic probability m ₁ (A _i ) is:

Wherein, n _i is the ith character in the training set, d _i is the Euclidean distance between the test sample and the ith character; α is 1 or 2;

The second basic probability distribution m ₂ (B _i ) is:

Among them, K is the total number of license plate types, k∈(0,K], W is the matrix composed of weights, W ^T is the transpose of W, b is the bias value of the multi-class logistic regression algorithm, and x _i is the i-th test samples, y _i is the license plate type of the ith test sample;

The third basic probability m ₃ (C _i ) distribution is:

Among them, n _i is the i-th vehicle color feature, _xi is the i-th test sample, and y _i is the i-th license plate color feature.

Wherein, the fusion rule is:

φ为辨识框架，s为所述辨识框架中所有支持车牌识别的证据集合。in,

φ is the identification frame, and s is the set of all evidence supporting license plate recognition in the identification frame.

To achieve the above purpose, the present invention provides a license plate detection system, comprising:

an extraction module for acquiring a test sample and extracting edge features, HOG features and HSV features of the test sample;

The first determination module is used to calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and determine the character corresponding to the minimum value of all the Euclidean distances as the target character;

The second determination module is used to obtain the license plate type through the training set, and use the multi-class logistic regression algorithm and the HOG feature to determine the target license plate type;

The third determination module is used to obtain the license plate color feature through the training set, construct an SVM classifier according to the license plate color feature, and input the HSV feature into the SVM classifier to obtain the target license plate color feature;

A detection module, configured to determine the license plate detection result of the test sample according to the target character, the type of the target license plate, and the color feature of the target license plate.

To achieve the above purpose, the present invention provides a license plate detection device, comprising:

memory for storing computer programs;

The processor is configured to implement the steps of the above-mentioned license plate detection method when executing the computer program.

To achieve the above object, the present invention provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the above-mentioned license plate detection method are implemented.

It can be seen from the above solutions that a license plate detection method provided by the present invention includes: acquiring a test sample, and extracting edge features, HOG features and HSV features of the test sample; The Euclidean distance of each character, and the character corresponding to the minimum value of all the Euclidean distances is determined as the target character; the license plate type is obtained through the training set, and the multi-class logistic regression algorithm and the HOG feature are used to determine the target license plate Type; obtain the license plate color feature through the training set, construct an SVM classifier according to the license plate color feature, and input the HSV feature into the SVM classifier to obtain the target license plate color feature; According to the target character, the target The license plate type and the color feature of the target license plate determine the license plate detection result of the test sample.

The license plate detection method provided by the present invention extracts the edge feature, HOG feature and HSV feature of the test sample, and adopts different classification methods for different features, so as to obtain the target character, target license plate type and target license plate color feature of the test sample. , and determine the final test result. Compared with the prior art scheme that detects a single feature and adopts a single classification method, the detection result is obtained by combining multiple features and multiple classification methods, and the integration of multiple data improves the multi-viewpoint and occlusion problems in multi-lane traffic scenarios. License plate detection accuracy. The invention also discloses a license plate detection system and device and a computer-readable storage medium, which can also achieve the above technical effects.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.

1 is a flowchart of a method for detecting a license plate disclosed in an embodiment of the present invention;

2 is a flowchart of another license plate detection method disclosed in an embodiment of the present invention;

3 is a structural diagram of a license plate detection system disclosed in an embodiment of the present invention;

FIG. 4 is a structural diagram of a license plate detection device disclosed in an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The embodiment of the invention discloses a license plate detection method, which improves the license plate detection accuracy.

Referring to FIG. 1, a flowchart of a license plate detection method disclosed in an embodiment of the present invention, as shown in FIG. 1, includes:

S101: Obtain a test sample, and extract edge features, HOG features, and HSV features of the test sample;

In the specific implementation, the training samples are actually vehicle images. The test samples obtained here can be collected by ordinary cameras, or by other cameras such as high-definition and ultra-high-definition cameras. Of course, they can also be collected by other devices capable of image collection. The acquired images are not specifically limited here. After obtaining the vehicle image, the image needs to be preprocessed to extract edge features, HOG (full name in Chinese: Histogram of oriented gradients, full name in English: Histogram of oriented gradients) features and HSV (full name in English: Hue, Saturation, Value, yes According to the intuitive characteristics of color, a color space created by A.R. Smith in 1978, also known as the hexagonal pyramid model) features, the preprocessing operations here include conventional technical means in the field of image processing.

S102: Calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and determine the character corresponding to the minimum value of all the Euclidean distances as the target character;

It can be understood that the edge features are the license plate characters in the vehicle picture. The test set contains multiple license plates, and each license plate contains multiple characters, such as A, B, Jing, Yue, etc. The Euclidean distance is used to represent the similarity between the license plate character and each character in the test set. The character with the highest similarity, that is, the corresponding character with the smallest Euclidean distance, is the target character of the license plate character.

S103: Obtain the license plate type through the training set, and use the multi-class logistic regression algorithm and the HOG feature to determine the target license plate type;

In a specific implementation, the license plates in the test set are used to obtain all types of license plates, such as 8-character new energy license plates, armed police license plates, coach license plates, Hong Kong and Macau license plates, etc. The HOG feature describes the local gradient magnitude and direction features of the image. HOG allows blocks to overlap each other, so it is not sensitive to illumination changes and small offsets, and can effectively describe edge features. Works well. The license plate type of the test sample is represented by the HOG feature, and the multi-class logistic regression algorithm is used to determine which license plate type the test sample belongs to, so as to ensure the best performance of the multi-class logistic regression and reduce the occurrence of overfitting during the training of the LR classifier. matching situation.

S104: Obtain the license plate color feature through the training set, construct an SVM classifier according to the license plate color feature, and input the HSV feature into the SVM classifier to obtain the target license plate color feature;

In a specific implementation, the license plates in the test set are used to obtain all kinds of license plate color features, such as white, blue, yellow, etc. The color features of the test samples are represented by HSV features, and the SVM (full name in English: Support VectorMachine, full name in Chinese: Support Vector Machine, is a fast pattern recognition method) classifier is used to determine which background color the license plate of the test sample belongs to.

It should be noted that, the above-mentioned execution processes of S102, S103, and S104 have no sequence.

S105: Determine the license plate detection result of the test sample according to the target character, the type of the target license plate, and the color feature of the target license plate.

The target character, target license plate type and target license plate color feature obtained in the above steps together constitute the license plate detection result of the test sample.

The license plate detection method provided by the embodiment of the present invention extracts the edge feature, HOG feature and HSV feature of the test sample, and adopts different classification methods for different features, so as to obtain the target character, target license plate type, target license plate of the test sample. color features and determine the final detection result. Compared with the prior art scheme that detects a single feature and adopts a single classification method, the detection result is obtained by combining multiple features and multiple classification methods, and the integration of multiple data improves the multi-viewpoint and occlusion problems in multi-lane traffic scenarios. License plate detection accuracy.

The embodiment of the present invention discloses a license plate detection method. Compared with the previous embodiment, this embodiment further describes and optimizes the technical solution. specific:

Referring to FIG. 2 , a flowchart of another license plate detection method provided by an embodiment of the present invention, as shown in FIG. 2 , includes:

S201: Obtain a test sample, and extract edge features, HOG features, and HSV features of the test sample;

S221: Calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and normalize all the Euclidean distances;

S222: Calculate the first basic probability distribution of the Euclidean distance after the normalization process, and determine the character corresponding to the maximum value among all the basic probabilities as the target character;

将其归一化范围为(0,1)，其中，j∈[1,i]。It can be understood that since the true value of the Euclidean distance is [0, ∞), the Euclidean distance needs to be normalized, that is

Normalize it to the range (0,1), where j∈[1,i].

The distribution of the first basic probability m ₁ (A _i ) is:

Wherein, n _i is the ith character in the training set, d _i is the Euclidean distance between the test sample and the ith character; α is 1 or 2;

That is to say, if the Euclidean distance is 0, the similarity is the largest. At this time, the confidence is 1, and the Euclidean distance is not 0, then the confidence corresponding to the character is calculated according to the above formula.

S203: Obtain the license plate type through the training set, use the multi-class logistic regression algorithm according to the HOG feature, calculate the second basic distribution probability of the license plate type, and determine the license plate type corresponding to the maximum value among all the basic probabilities as the target license plate type;

The second basic probability distribution m ₂ (B _i ) is:

Among them, K is the total number of license plate types, k∈(0,K], W is the matrix composed of weights, W ^T is the transpose of W, b is the bias value of the multi-class logistic regression algorithm, and x _i is the i-th test samples, y _i is the license plate type of the ith test sample;

S204: Obtain the license plate color feature through the training set, use the SVM classifier according to the HSV feature, calculate the third basic probability distribution of the license plate color feature, and assign the license plate color corresponding to the maximum value of all the basic probabilities The feature is determined as the color feature of the target license plate;

The third basic probability m ₃ (C _i ) distribution is:

Among them, n _i is the i-th vehicle color feature, _xi is the i-th test sample, and y _i is the i-th license plate color feature.

In a specific implementation, the SVM classifier uses the Sigmod function as the connection function, and the output range is 0-1. The SVM classifier can use the above formula to calculate the probability distribution of the license plate color feature.

S205: Determine the license plate detection result of the test sample according to the target character, the target license plate type, and the color feature of the target license plate;

S206: Calculate the degree of confidence of the license plate detection result according to the first basic probability distribution, the second basic probability distribution, and the third basic probability distribution according to a preset fusion rule.

On the identification frame θ, there are three mutually independent evidences A _i , B _i , C _i , and their basic probability assignment functions are m ₁ (A _i ), m ₂ (B _i ), m ₃ (C _i respectively ), those skilled in the art can select suitable fusion rules to fuse according to the actual situation, which is not specifically limited here, preferably:

φ为辨识框架，s为所述辨识框架中所有支持车牌识别的证据集合。in,

φ is the identification frame, and s is the set of all evidence supporting license plate recognition in the identification frame.

A license plate detection system provided by an embodiment of the present invention is introduced below. A license plate detection system described below and a license plate detection method described above can be referred to each other.

Referring to FIG. 3, a structural diagram of a license plate detection system provided by an embodiment of the present invention, as shown in FIG. 3, includes:

The extraction module 301 is used to obtain a test sample, and extract the edge feature, HOG feature and HSV feature of the test sample;

The first determination module 302 is used to calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and determine the character corresponding to the minimum value of all the Euclidean distances as the target character;

The second determination module 303 is used to obtain the license plate type through the training set, and determine the target license plate type by using the multi-class logistic regression algorithm and the HOG feature;

The third determination module 304 is used to obtain the license plate color feature through the training set, construct an SVM classifier according to the license plate color feature, and input the HSV feature into the SVM classifier to obtain the target license plate color feature;

The detection module 305 is configured to determine the license plate detection result of the test sample according to the target character, the type of the target license plate, and the color feature of the target license plate.

The license plate detection system provided by the embodiment of the present invention extracts edge features, HOG features and HSV features of the test sample, and adopts different classification methods for different features, so as to obtain the target character, target license plate type, target license plate of the test sample. color features and determine the final detection result. Compared with the prior art scheme that detects a single feature and adopts a single classification method, the detection result is obtained by combining multiple features and multiple classification methods, and the integration of multiple data improves the multi-viewpoint and occlusion problems in multi-lane traffic scenarios. License plate detection accuracy.

On the basis of the foregoing embodiment, as a preferred implementation manner, the first determining module 302 includes:

A normalization unit, used for calculating the Euclidean distance between the test sample and each character in the training set according to the edge feature, and normalizing all the Euclidean distances;

The determining unit is configured to calculate the first basic probability distribution of the Euclidean distance after the normalization process, and determine the character corresponding to the maximum value among all the basic probabilities as the target character.

On the basis of the above embodiment, as a preferred embodiment, the second determining module 303 specifically obtains the license plate type through the training set, and uses a multi-class logistic regression algorithm according to the HOG feature to calculate the second The basic distribution probability is determined, and the license plate type corresponding to the maximum value of all the basic probabilities is determined as the module of the target license plate type.

On the basis of the above embodiment, as a preferred implementation manner, the third determination module 304 includes:

A construction unit for obtaining license plate color features through the training set, and constructing an SVM classifier according to the license plate color features;

A computing unit, configured to utilize the SVM classifier according to the HSV feature, calculate the third basic probability distribution of the license plate color feature, and determine the license plate color feature corresponding to the maximum value in all basic probabilities as the target license plate color characteristics.

On the basis of the above embodiment, as a preferred embodiment, it also includes:

A calculation module, configured to calculate the trust degree of the license plate detection result according to the first basic probability distribution, the second basic probability distribution and the third basic probability distribution according to a preset fusion rule.

On the basis of the above embodiment, as a preferred embodiment, the distribution of the first basic probability m ₁ (A _i ) is:

Wherein, n _i is the ith character in the training set, d _i is the Euclidean distance between the test sample and the ith character; α is 1 or 2;

The second basic probability distribution m ₂ (B _i ) is:

Among them, K is the total number of license plate types, k∈(0,K], W is the matrix composed of weights, W ^T is the transpose of W, b is the bias value of the multi-class logistic regression algorithm, and x _i is the i-th test samples, y _i is the license plate type of the ith test sample;

The third basic probability m ₃ (C _i ) distribution is:

Among them, n _i is the i-th vehicle color feature, _xi is the i-th test sample, and y _i is the i-th license plate color feature.

On the basis of the above embodiment, as a preferred implementation, the fusion rule is:

φ为辨识框架，s为所述辨识框架中所有支持车牌识别的证据集合。in,

φ is the identification frame, and s is the set of all evidence supporting license plate recognition in the identification frame.

The application also provides a license plate detection device. Referring to FIG. 4 , a structural diagram of a license plate detection device provided by an embodiment of the present invention, as shown in FIG. 4 , includes:

memory 401 for storing computer programs;

The processor 402 can implement the steps provided in the above embodiments when executing the computer program. Of course, the license plate detection device may also include various network interfaces, power supplies and other components.

The license plate detection device provided by the embodiment of the present invention extracts edge features, HOG features and HSV features of the test sample, and adopts different classification methods for different features, so as to obtain the target character, target license plate type, target license plate of the test sample. color features and determine the final detection result. Compared with the solution in the prior art in which a single classification method is adopted to detect a single feature, the accuracy of license plate detection is improved. It can be seen that the license plate detection device provided by the embodiment of the present invention obtains the detection result by combining multiple features and multiple classification methods, and the integration of multiple data improves the license plate detection accuracy of the multi-viewpoint and occlusion problems in the multi-lane traffic mode scene .

The present application also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps provided in the above embodiments can be implemented. The storage medium may include: U disk, removable hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other media that can store program codes.

The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the various embodiments can be referred to each other.

The above description of the disclosed embodiments enables any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the various embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant part can be referred to the description of the method. It should be pointed out that for those of ordinary skill in the art, without departing from the principles of the present application, several improvements and modifications can also be made to the present application, and these improvements and modifications also fall within the protection scope of the claims of the present application.

It should also be noted that, in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply these entities or operations. There is no such actual relationship or sequence between operations. Moreover, the terms "comprising", "comprising" or any other variation thereof are intended to encompass non-exclusive inclusion such that a process, method, article or device comprising a list of elements includes not only those elements, but also includes not explicitly listed or other elements inherent to such a process, method, article or apparatus. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in a process, method, article or apparatus that includes the element.

Claims (9)

1. a license plate detection method, is characterized in that, comprises: Obtain a test sample, and extract the edge feature, HOG feature and HSV feature of the test sample; Calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and determine the character corresponding to the minimum value of all the Euclidean distances as the target character; Obtain the license plate type through the training set, and use the multi-class logistic regression algorithm and the HOG feature to determine the target license plate type; Obtain the license plate color feature through the training set, construct an SVM classifier according to the license plate color feature, and input the HSV feature into the SVM classifier to obtain the target license plate color feature; Determine the license plate detection result of the test sample according to the target character, the target license plate type, and the color feature of the target license plate; Wherein, calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and determine the character corresponding to the minimum value of all the Euclidean distances as the target character, including: Calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and normalize all the Euclidean distances; Calculate the first basic probability distribution of the Euclidean distance after the normalization process, and determine the character corresponding to the maximum value among all basic probabilities as the target character; Wherein, the distribution of the first basic probability m ₁ (A _i ) is:

Wherein, n _i is the ith character in the training set, p(i) is the basic probability of the ith character, d _i is the Euclidean distance between the test sample and the ith character; α is 1 or 2. 2. license plate detection method according to claim 1, is characterized in that, utilizes the logistic regression algorithm of multi-classification and described HOG feature to determine target license plate type, comprising: According to the HOG feature, the multi-class logistic regression algorithm is used to calculate the second basic probability distribution of the license plate type, and the license plate type corresponding to the maximum value among all the basic probabilities is determined as the target license plate type. 3. according to the described license plate detection method of claim 2, it is characterised in that described HSV feature is inputted in the described SVM classifier to obtain target license plate color feature, comprising: Using the SVM classifier according to the HSV feature, the third basic probability distribution of the license plate color feature is calculated, and the license plate color feature corresponding to the maximum value among all the basic probabilities is determined as the target license plate color feature. 4. license plate detection method according to claim 3, is characterized in that, also comprises: The trust degree of the license plate detection result is calculated according to the first basic probability distribution, the second basic probability distribution and the third basic probability distribution according to a preset fusion rule. 5. The license plate detection method according to claim 4, wherein the second basic probability distribution m ₂ (B _i ) is:

Among them, K is the total number of license plate types, k∈(0,K], W _k is the matrix composed of the weights corresponding to the k-th license plate type,

is the transpose of W _k , b _k is the bias value of the multi-class logistic regression algorithm corresponding to the k th license plate type, x _i is the ith test sample, and y _i is the license plate type of the ith test sample; The third basic probability m ₃ (C _i ) distribution is:

Among them, n _i is the ith vehicle color feature, x is the test sample, y is the license plate color feature, _xi is the ith test sample, and _yi is the ith license plate color feature. 6. license plate detection method according to claim 5, is characterized in that, described fusion rule is:

in,

φ is the identification frame, and s is the set of all evidence supporting license plate recognition in the identification frame. 7. A license plate detection system, comprising: an extraction module for acquiring a test sample and extracting edge features, HOG features and HSV features of the test sample; The first determination module is used to calculate the Euclidean distance between the test sample and each character in the training set according to the edge feature, and determine the character corresponding to the minimum value of all the Euclidean distances as the target character; The second determination module is used to obtain the license plate type through the training set, and use the multi-class logistic regression algorithm and the HOG feature to determine the target license plate type; The third determination module is used to obtain the license plate color feature through the training set, construct an SVM classifier according to the license plate color feature, and input the HSV feature into the SVM classifier to obtain the target license plate color feature; a detection module, configured to determine the license plate detection result of the test sample according to the target character, the type of the target license plate, and the color feature of the target license plate; Wherein, the first determining module includes: A normalization unit, used for calculating the Euclidean distance between the test sample and each character in the training set according to the edge feature, and normalizing all the Euclidean distances; a determining unit, used for calculating the first basic probability distribution of the Euclidean distance after normalization, and determining the character corresponding to the maximum value in all basic probabilities as the target character; Wherein, the distribution of the first basic probability m ₁ (A _i ) is:

Wherein, n _i is the ith character in the training set, p(i) is the basic probability of the ith character, d _i is the Euclidean distance between the test sample and the ith character; α is 1 or 2. 8. A license plate detection device, comprising: memory for storing computer programs; The processor is configured to implement the steps of the license plate detection method according to any one of claims 1 to 6 when executing the computer program. 9. A computer-readable storage medium, characterized in that, a computer program is stored on the computer-readable storage medium, and when the computer program is executed by the processor, the license plate detection as described in any one of claims 1 to 6 is realized steps of the method.

Images