CN110929589A - Method, device, computer device and storage medium for vehicle feature recognition - Google Patents
Method, device, computer device and storage medium for vehicle feature recognition Download PDFInfo
- Publication number
- CN110929589A CN110929589A CN201911050662.9A CN201911050662A CN110929589A CN 110929589 A CN110929589 A CN 110929589A CN 201911050662 A CN201911050662 A CN 201911050662A CN 110929589 A CN110929589 A CN 110929589A
- Authority
- CN
- China
- Prior art keywords
- image
- vehicle
- recognition
- license plate
- feature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/54—Surveillance or monitoring of activities, e.g. for recognising suspicious objects of traffic, e.g. cars on the road, trains or boats
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/63—Scene text, e.g. street names
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/625—License plates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Multimedia (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method, a device, a computer device and a storage medium for vehicle feature recognition, wherein a first recognition image of a vehicle is acquired, processing the preset parameters of the first identification image through an image enhancement model to obtain a second identification image, the geographical position information of the vehicle is obtained through the position recognition model, the second recognition image is recognized through the vehicle cover detection model, the second recognition image is recognized through the license plate detection model, under the condition that the detection result of the cover of the second identification image is not strict or the detection result of the license plate is absent, the second identification image is marked as unsafe, the geographic position information of the vehicle is uploaded and a warning signal is sent out, so that under the condition of insufficient light at night, the problem of low recognition degree of the muck vehicle features is solved, and the efficiency and the accuracy of vehicle night image feature recognition are improved.
Description
Technical Field
The present application relates to the field of image recognition technology, and in particular, to a method, device, computer device, and storage medium for vehicle feature recognition.
Background
With the rapid advance of the urbanization process in China, infrastructure of various regions is vigorous, and the earth-moving vehicles are used more and more as main force for transporting infrastructure materials and other construction wastes. However, the muck truck generally transports capital construction materials and other construction wastes at night, if the cover of the muck truck is not tightly shielded, dust is raised along the road in the whole driving process of the muck truck, the phenomenon of remaining muck occurs, and serious pollution is caused to the urban environment; moreover, the muck vehicle has large vehicle type, high driving speed at night and relatively small license plate, so that the muck vehicle is not beneficial to being quickly positioned and identified, and if an accident occurs, the muck vehicle is easy to escape, so that the muck vehicle is required to be written with an enlarged license plate on a vehicle cover, the enlarged license plate can be formed by printing or handwriting, but the muck vehicle is easy to stain and shield the enlarged license plate due to long-term transportation of muck. In addition, insufficient light at night can also affect the feature recognition of the muck vehicle.
In the related technology, the two-way camera is connected, the head and the tail of the vehicle are simultaneously captured and matched to identify the muck vehicle, whether the muck vehicle has a behavior violating traffic rules or not is identified, however, under the condition of insufficient light at night, the picture acquired by the camera is fuzzy, and the method has low identification degree on the characteristics of whether the cover of the muck vehicle is tightly shielded, whether the muck vehicle has a complete amplification license plate or not and the like.
Aiming at the problem that the recognition degree of the characteristics of the muck vehicle is low under the condition of insufficient light at night in the related technology, an effective solution is not provided at present.
Disclosure of Invention
The invention provides a method, equipment, computer equipment and a storage medium for identifying vehicle characteristics, aiming at the problem that the identification degree of the characteristics of a muck vehicle is low under the condition of insufficient light at night in the related art, and aims to at least solve the problem.
According to an aspect of the present invention, there is provided a method of vehicle feature identification, the method comprising:
acquiring a first identification image of a vehicle, and processing preset parameters of the first identification image through an image enhancement model to obtain a second identification image;
acquiring geographic position information of a vehicle through a position recognition model, recognizing the second recognition image through a cover detection model, judging whether the cover of the vehicle in the second recognition image is covered tightly, recognizing the second recognition image through a license plate detection model, and judging whether the vehicle in the second recognition image has an amplified license plate, wherein the amplified license plate has complete and identifiable license plate information including the amplified license plate;
and under the condition that the hood detection result of the second identification image is not strict or the license plate detection result is absent, marking the second identification image as unsafe, uploading the geographic position information of the vehicle and sending out a warning signal.
In one embodiment, before the identifying the second identification image by the hood detection model, the method further includes:
segmenting and screening the second identification image through a target detection network algorithm to obtain a vehicle image in the second identification image and an image confidence coefficient of the vehicle image, and discarding the vehicle image corresponding to the image confidence coefficient under the condition that the image confidence coefficient is smaller than a first preset confidence coefficient;
under the condition that the image confidence coefficient is smaller than a second preset confidence coefficient and is larger than or equal to the first preset confidence coefficient, carrying out false detection analysis on the vehicle image through a convolutional neural network and a classifier algorithm, and judging whether the vehicle image is the image of the target vehicle or not according to the judgment result of the classifier algorithm;
and if the vehicle image is not the target vehicle, discarding the vehicle image corresponding to the image confidence.
In one embodiment, the identifying the second identification image by the hood detection model includes:
extracting a first gradient feature and a first convolution neural network feature of the second identification image;
reducing the dimension of the first gradient feature to obtain a second gradient feature, and reducing the dimension of the first convolution neural network feature to obtain a second convolution neural network feature;
fusing the second gradient feature and the second convolutional neural network feature to obtain a fused feature;
and processing the fusion features through a linear regression model to obtain a fitting coefficient of the linear regression model, judging that the detection result of the vehicle cover is tight under the condition that the fitting coefficient is greater than or equal to a preset coefficient value, and judging that the detection result of the vehicle cover is not tight under the condition that the fitting coefficient is less than the preset coefficient value.
In one embodiment, the recognizing the second recognition image by the license plate detection model includes:
acquiring characters in the second recognition image to obtain the number of the characters and the confidence coefficient of the characters;
and judging that the license plate detection result exists when the number of the characters is greater than or equal to a preset number and the character confidence of each character is greater than or equal to a third preset confidence, and judging that the license plate detection result does not exist when the number of the characters is less than the preset number or the character confidence is less than the third preset confidence.
In one embodiment, the obtaining the vehicle geographic position information through the position identification model includes:
and acquiring the geographical position information of the monitoring equipment which captures the first identification image as the geographical position information of the vehicle.
In one embodiment, the processing the preset parameter of the first recognition image through the image enhancement model to obtain the second recognition image includes:
and carrying out contrast and brightness processing on the first identification image through an image enhancement model to obtain a second identification image, wherein the contrast and brightness processing comprises enhancing the contrast of the first identification image and improving the brightness of the first identification image.
In one embodiment, after the identifying the second identification image by the hood detection model, the method further comprises:
and under the conditions that the hood detection result of the second identification image is strict and the license plate detection result exists, marking the second identification image as safe.
According to another aspect of the present invention, there is provided an apparatus for vehicle feature recognition, the apparatus comprising an enhancement module, a detection module, and a determination module:
the enhancement module is used for acquiring a first identification image of a vehicle and processing the preset parameters of the first identification image through an image enhancement model to obtain a second identification image;
the detection module is used for acquiring vehicle geographic position information through a position recognition model, recognizing the second recognition image through a cover detection model, judging whether the cover of the vehicle in the second recognition image is tightly covered or not, recognizing the second recognition image through a license plate detection model, and judging whether an amplified license plate exists in the vehicle in the second recognition image or not, wherein the existence of the amplified license plate comprises the complete and identifiable license plate information of the amplified license plate;
the judging module is configured to mark the second recognition image as unsafe, upload the geographic position information of the vehicle, and issue a warning signal when the hood detection result of the second recognition image is not strict or the license plate detection result of the second recognition image is absent.
In one embodiment, the apparatus further comprises a processing module, a false detection module, and a screening module:
the processing module is used for segmenting and screening the second identification image through a target detection network algorithm to obtain a vehicle image in the second identification image and an image confidence coefficient of the vehicle image, and the vehicle image corresponding to the image confidence coefficient is discarded under the condition that the image confidence coefficient is smaller than a first preset confidence coefficient;
the false detection module is used for carrying out false detection analysis on the vehicle image through a convolutional neural network and a classifier algorithm under the condition that the image confidence coefficient is smaller than a second preset confidence coefficient and is larger than or equal to the first preset confidence coefficient, and judging whether the vehicle image is the image of the target vehicle or not according to the judgment result of the classifier algorithm;
the screening module is used for discarding the vehicle image corresponding to the image confidence coefficient under the condition that the vehicle image is not the target vehicle.
The extraction unit is used for extracting a first gradient feature and a first convolution neural network feature of the second identification image;
the dimension reduction unit is used for reducing the dimension of the first gradient feature to obtain a second gradient feature, and reducing the dimension of the first convolutional neural network feature to obtain a second convolutional neural network feature;
the fusion unit is used for fusing the second gradient feature and the second convolutional neural network feature to obtain a fusion feature;
the judging unit is used for processing the fusion features through a linear regression model to obtain a fitting coefficient of the linear regression model, judging that the detection result of the vehicle cover is tight under the condition that the fitting coefficient is greater than or equal to a preset coefficient value, and judging that the detection result of the vehicle cover is not tight under the condition that the fitting coefficient is less than the preset coefficient value.
According to another aspect of the present invention, there is provided a computer device comprising a memory storing a computer program and a processor implementing any of the methods described above when the processor executes the computer program.
According to another aspect of the invention, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements any of the methods described above.
According to the invention, a first identification image of a vehicle is obtained, the preset parameters of the first identification image are processed through an image enhancement model to obtain a second identification image, the geographical position information of the vehicle is obtained through a position identification model, the second identification image is identified through a cover detection model to judge whether the cover of the vehicle in the second identification image is covered tightly, the second identification image is identified through a license plate detection model to judge whether the vehicle in the second identification image has an amplified license plate, wherein the amplified license plate with the amplified license plate has complete and identifiable license plate information, the second identification image is marked as unsafe under the condition that the cover detection result of the second identification image is not tight or the license plate detection result is not existed, the geographical position information of the vehicle is uploaded and a warning signal is sent out, the problem of under the not enough condition of light at night, to the lower problem of degree of recognition of dregs car characteristic is solved, efficiency and the degree of accuracy to vehicle image feature identification at night have been improved.
Drawings
FIG. 1 is a schematic diagram of an application environment of a method for vehicle feature recognition according to an embodiment of the invention;
FIG. 2 is a first flowchart of a method of vehicle feature identification according to an embodiment of the present invention;
FIG. 3 is a second flowchart of a method of vehicle feature identification according to an embodiment of the present invention;
FIG. 4 is a flow chart of a method of hood detection according to an embodiment of the invention;
FIG. 5 is a flow chart of a method of license plate detection according to an embodiment of the present invention;
FIG. 6 is a first block diagram of the vehicle feature recognition apparatus according to an embodiment of the present invention;
FIG. 7 is a block diagram II of the vehicle feature recognition apparatus according to the embodiment of the present invention;
fig. 8 is a block diagram of the structure of an apparatus for vehicle feature recognition according to an embodiment of the present invention;
FIG. 9 is a schematic diagram of a vehicle feature recognition system according to an embodiment of the present invention;
fig. 10 is a flowchart three of a method of vehicle feature identification according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The method for remotely controlling a vehicle provided by the present application may be applied to an application environment shown in fig. 1, and fig. 1 is an application environment schematic diagram of a method for identifying vehicle features according to an embodiment of the present invention, as shown in fig. 1, where a server 104 obtains a first identification image from a terminal 102, performs image enhancement processing and feature identification on the first identification image, obtains image features of a vehicle in the first identification image, and determines the image features of the vehicle, where the server 104 may be implemented by an independent server or a server cluster composed of a plurality of servers, and the terminal 102 may be a monitoring device, and may also be a personal computer, a notebook computer, a smart phone, a tablet computer, and a portable wearable device.
In one embodiment, a method for vehicle feature recognition is provided, and fig. 2 is a flowchart illustrating a method for vehicle feature recognition according to an embodiment of the present invention, as shown in fig. 2, the method includes the following steps:
step S202, a first identification image of a vehicle is obtained, and preset parameters of the first identification image are processed through an image enhancement model to obtain a second identification image, wherein the vehicle in the first identification image can be a muck truck, a van and the like, the first identification image of the vehicle can be obtained through monitoring equipment installed on a street or a toll station, can be obtained through video data of a monitoring platform of a traffic department, and can also be obtained through mobile equipment, the mobile equipment comprises a mobile terminal and the like with a photographing or shooting function, a user can also upload the first identification image to a server for processing and identification, and the preset parameters can be chromaticity, contrast, sharpening degree and the like. Because the light is insufficient at night, the server needs to perform image enhancement processing on the first identification image to obtain a second identification image, and the image enhancement model can be realized by an LECARM algorithm, so that the boundary between the vehicle and the environment background is clearer, and the feature extraction of the second identification image is facilitated.
Step S204, obtaining vehicle geographic position information through a position recognition model, recognizing the second recognition image through a cover detection model, judging whether the cover of the vehicle in the second recognition image is covered tightly, and judging whether the vehicle in the second recognition image has a magnified license plate through a license plate detection model, wherein when the position recognition model obtains the first recognition image shot by the mobile device, a positioning System of the mobile device provides geographic position information as the vehicle geographic position information of the second recognition image, the geographic position information can be a bayonet with a checking facility, an electric alarm scene with electronic equipment, a city street and the like, the positioning System comprises a Global Positioning System (GPS) and a Beidou satellite Navigation System (BDS), and a user can also upload the geographic position when obtaining the first recognition image, and matching the license plate with a license plate database in a public security system, and considering that the amplified license plate is clear, complete and identifiable under the condition of successful matching.
Step S206, when the hood detection result of the second recognition image is not strict or the license plate detection result is absent, the second recognition image is marked as unsafe, the geographic location information of the vehicle is uploaded and a warning signal is issued, and the server may further contact the police to handle the offending vehicle when the second recognition image is marked as unsafe.
Through the steps S202 to S206, the server acquires the first identification image from the terminal, the second identification image is obtained through image enhancement, the second identification image is subjected to vehicle cover and amplified license plate identification, whether warning is given according to the identification result or not is judged, the image enhancement model enables the image acquired at night to be processed and identified more easily, the problem that the identification degree of the muck vehicle characteristics is low under the condition that light is insufficient at night is solved, and the efficiency and the accuracy of vehicle night image characteristic identification are improved.
In one embodiment, fig. 3 is a flowchart of a method for vehicle feature identification according to an embodiment of the present invention, and as shown in fig. 3, the method further includes the following steps:
step S302, segmenting and screening the second identification image through a target detection network algorithm to obtain a vehicle image in the second identification image and an image confidence coefficient of the vehicle image, discarding the vehicle image corresponding to the image confidence coefficient when the image confidence coefficient is smaller than a first preset confidence coefficient, wherein a target detection network algorithm (You Only LookOne, which is abbreviated as YoLO) adopts an algorithm of a YoLO V3 version to judge whether the vehicle in the second identification image is a target vehicle, the target vehicle is a vehicle to be identified, such as a muck truck, a van and the like, the YoLO V3 identifies position coordinates of the vehicle in the second identification image, segmenting the second identification image according to the position coordinates, Only reserving the vehicle image in the second identification image, discarding a background environment of the vehicle, and acquiring the image confidence coefficient of the vehicle image, for example, in the case where the target vehicle is a muck truck and the first preset confidence is 20, the vehicle image with the image confidence less than 20 is discarded and is considered not to be a muck truck.
Step S304, when the image confidence is smaller than the second preset confidence and is greater than or equal to the first preset confidence, performing false detection analysis on the vehicle image through a Convolutional Neural network algorithm and a classifier algorithm, judging whether the vehicle image is the image of the target vehicle according to the judgment result of the classifier algorithm, considering the vehicle corresponding to the image confidence greater than or equal to 85 as the target vehicle when the first preset confidence is 20 and the second preset confidence is 85, performing false detection analysis on the vehicle with the image confidence greater than or equal to 20 and the image confidence less than 85, analyzing the vehicle image through a Convolutional Neural Network (CNN) and the classifier algorithm, wherein the classifier algorithm can adopt a Support Vector Machine algorithm (SVM), SVM models are trained through a large amount of SVM, and analyzing the vehicle image needing to be identified, and judging whether the vehicle image is the target vehicle.
In step S306, if the vehicle image is not the target vehicle, the vehicle image corresponding to the image confidence is discarded.
Through the steps S302 to S306, the second identification image is screened and analyzed by adopting the target detection network algorithm, the convolutional neural network algorithm and the classifier algorithm, and the vehicle image which is not the target vehicle in the second identification image is discarded, so that the calculated amount in the subsequent image processing process is reduced, and the efficiency of image feature identification is improved.
In one embodiment, fig. 4 is a flowchart of a method of vehicle hood detection according to an embodiment of the invention, as shown in fig. 4, the method comprising the steps of:
step S402, extracting a first Gradient feature and a first convolution neural network feature of the second recognition image, where the extracting of the first Gradient feature may be implemented by a Histogram of Oriented Gradients (HOG) algorithm, and extracting edge Gradient information of the second recognition image by using the HOG algorithm to perform feature extraction.
Step S404, performing dimensionality reduction on the first gradient feature to obtain a second gradient feature, performing dimensionality reduction on the first convolutional neural network feature to obtain a second convolutional neural network feature, where the dimensionality extraction is high and the data processing amount is large, the dimensionality reduction needs to be performed on the first gradient feature and the first convolutional neural network feature, a dimensionality reduction algorithm may use a Principal Component Analysis (PCA) method, and the PCA algorithm maps the original dimensionality feature in a new dimensionality coordinate, ignores the dimensionality feature whose variance is almost 0 in the new dimensionality coordinate, and implements dimensionality reduction processing on the data feature.
And step S406, fusing the second gradient feature and the second convolutional neural network feature to obtain a fused feature, wherein in the feature fusion process, the second gradient feature and the second convolutional neural network feature can be simply stacked to realize fusion, and the second gradient feature and the second convolutional neural network feature can also be linearly stacked and fused through a functional relationship.
Step S408, processing the fusion features through a linear regression model to obtain a fitting coefficient of the linear regression model, judging that the detection result of the vehicle cover is strict when the fitting coefficient is greater than or equal to a preset coefficient value, and judging that the detection result of the vehicle cover is not strict when the fitting coefficient is less than the preset coefficient value, wherein the value range of the fitting coefficient is greater than or equal to 0 and less than or equal to 1, the closer the fitting coefficient is to 1, the better the fitting effect is, for example, the preset coefficient value is 0.95, and the strict detection result of the vehicle cover is shown when the fitting coefficient of the linear regression model is 0.96.
Through the steps S402 to S408, the image features in the second identification image are extracted, reduced in dimension and fused, so that the main features are kept, the data processing amount is reduced, and the image feature identification efficiency is improved.
In one embodiment, fig. 5 is a flowchart of a license plate detection method according to an embodiment of the present invention, and as shown in fig. 5, the method includes the following steps:
step S502, obtaining characters in the second recognition image to obtain the number of the characters and the character confidence coefficient, wherein the characters in the second recognition image may be shielded or stained, so that all the characters in the second recognition image need to be detected through a license plate detection model, wherein the characters comprise Chinese characters, numbers and English words, the license plate detection model also obtains the character confidence coefficient corresponding to the characters when the characters are recognized, the character confidence coefficient represents the accuracy degree of character recognition, the value range of the character confidence coefficient is greater than or equal to 0 and less than or equal to 1, and the closer the character confidence coefficient is to 1, the higher the accuracy degree of character recognition is;
step S504, judging that the license plate detection result is present when the number of the characters is greater than or equal to a preset number and the character confidence of each character is greater than or equal to a third preset confidence, judging that the license plate detection result is absent when the number of the characters is less than the preset number or the character confidence is less than the third preset confidence, for example, the preset number is 3, the third character confidence is 0.8, and the license plate detection result is present when the number of the characters detected by the license plate detection model is 4 and the character confidence of each character is greater than or equal to 0.8; under the condition that the number of the detected characters is 4 and the confidence coefficient of one character is 0.75, the license plate detection result is absent; and under the condition that the number of the detected characters is 2 and the character confidence coefficient of each character is greater than or equal to 0.8, the license plate detection result still does not exist.
Through the steps S502 and S504, the license plate detection model determines whether the characters in the second recognition image are the amplified license plate meeting the conditions by detecting the characters in the second recognition image and obtaining the character confidence corresponding to the characters, so as to meet the requirements for vehicle feature recognition.
In one embodiment, the geographical location information of the monitoring device can be acquired and used as the geographical location information of the vehicle, wherein the monitoring device captures the first identification image, the monitoring device uploads the geographical location information of the monitoring device simultaneously in the process of uploading the images to the server, the geographical location information is used as the geographical location information of all the first identification images uploaded by the monitoring device, the accuracy of the geographical location information is improved, and the geographical location information is uploaded to the police under the condition that the detection result of the vehicle cover is not strict or the detection result of the vehicle plate does not exist, so that the management of the police on the illegal vehicle is facilitated.
In one embodiment, contrast and brightness processing is carried out on the first recognition image through an image enhancement model to obtain a second recognition image, wherein the contrast and brightness processing comprises enhancing the contrast of the first recognition image and improving the brightness of the first recognition image, the brightness of the first recognition image is improved through the image enhancement model, the contrast of the first recognition image is enhanced, and the efficiency and the accuracy of vehicle night image feature recognition are improved.
In one embodiment, after the second recognition image is recognized by the hood detection model, the method further includes marking the second recognition image as safe under the condition that the hood detection result of the second recognition image is strict and the license plate detection result is present, and under the condition that the second recognition image is marked as safe, the method does not perform any processing on the vehicle corresponding to the second recognition image, thereby simplifying the processing flow.
It should be understood that, although the steps in the flowcharts of fig. 2 to 5 are shown in sequence as indicated by the arrows, the steps are not necessarily performed in sequence as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 2-5 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performing the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.
Corresponding to the above method for identifying vehicle features, in this embodiment, a device for identifying vehicle features is further provided, and the device is used to implement the above embodiments and preferred embodiments, and is already described and is not repeated. As used below, the term "module" may be a combination of software and/or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementations in hardware or a combination of software and hardware are also possible and contemplated.
In one embodiment, an apparatus for vehicle feature recognition is provided, and fig. 6 is a block diagram illustrating a structure of the apparatus for vehicle feature recognition according to the embodiment of the present invention, as shown in fig. 6, including: an enhancement module 62, a detection module 64, and a decision module 66, wherein:
the enhancement module 62 is configured to obtain a first identification image of a vehicle, and process a preset parameter of the first identification image through an image enhancement model to obtain a second identification image;
the detection module 64 is configured to obtain geographic position information of the vehicle through the position recognition model, recognize the second recognition image through the cover detection model, determine whether a cover of the vehicle in the second recognition image is covered tightly, recognize the second recognition image through the license plate detection model, and determine whether an amplified license plate exists in the vehicle in the second recognition image, where the existence of the amplified license plate includes complete and identifiable license plate information of the amplified license plate;
the determination module 66 is configured to mark the second recognition image as unsafe if the hood detection result of the second recognition image is not strict or the license plate detection result of the second recognition image is absent, upload the geographic location information of the vehicle, and issue a warning signal.
Through the equipment, the server acquires a first identification image from the terminal, a second identification image is obtained through the enhancement module 62, the detection module 64 identifies the second identification image by means of the automobile cover and the amplified license plate, the judgment of whether warning is given by the judgment module 66 according to the identification result, the image acquired at night is easier to process and identify by the image enhancement module, the problem that the identification degree of the muck vehicle characteristics is low under the condition that light is insufficient at night is solved, and the efficiency and the accuracy of vehicle night image characteristic identification are improved.
In one embodiment, fig. 7 is a block diagram of a second structure of the apparatus for vehicle feature recognition according to the embodiment of the present invention, as shown in fig. 7, further including: a processing module 72, a false detection module 74, and a screening module 76, wherein:
the processing module 72 is configured to segment and filter the second identification image through a target detection network algorithm to obtain a vehicle image in the second identification image and an image confidence of the vehicle image, and discard the vehicle image corresponding to the image confidence when the image confidence is smaller than a first preset confidence;
the false detection module 74 is configured to, when the image confidence is smaller than the second preset confidence and greater than or equal to the first preset confidence, perform false detection analysis on the vehicle image through a convolutional neural network and a classifier algorithm, and determine whether the vehicle image is an image of a target vehicle according to a determination result of the classifier algorithm;
and the screening module 76 is used for discarding the vehicle image corresponding to the image confidence coefficient under the condition that the vehicle image is not the target vehicle.
Through the device, the processing module 72 adopts a target detection network algorithm, the false detection module 74 adopts a convolutional neural network algorithm and a classifier algorithm to screen and analyze the second identification image, and the screening module 76 discards the vehicle image which is not the target vehicle in the second identification image, so that the calculated amount in the subsequent image processing process is reduced, and the efficiency of image feature identification is improved.
In one embodiment, fig. 8 is a block diagram of a third structure of the apparatus for vehicle feature recognition according to the embodiment of the present invention, and as shown in fig. 8, the detection module 64 of the present invention includes: extraction unit 82, dimension reduction unit 84, fusion unit 86, and determination unit 88:
an extracting unit 82, configured to extract a first gradient feature and a first convolutional neural network feature of the second recognition image;
a dimension reduction unit 84, configured to perform dimension reduction on the first gradient feature to obtain a second gradient feature, and perform dimension reduction on the first convolutional neural network feature to obtain a second convolutional neural network feature;
a fusion unit 86, configured to fuse the second gradient feature and the second convolutional neural network feature to obtain a fusion feature;
the judging unit 88 is configured to process the fusion features through a linear regression model to obtain a fitting coefficient of the linear regression model, judge that the vehicle cover detection result is strict when the fitting coefficient is greater than or equal to a preset coefficient value, and judge that the vehicle cover detection result is not strict when the fitting coefficient is less than the preset coefficient value.
Through the device, the extracting unit 82 extracts the image features in the second recognition image, the dimension reducing unit 84 reduces the dimension of the extracted features, and the fusing unit 86 fuses the image features after dimension reduction, so that the main features are kept, the data processing amount is reduced, and the image feature recognition efficiency is improved.
The following describes an embodiment of the present invention in detail with reference to an actual application scenario, and when performing vehicle feature recognition, fig. 9 is a schematic diagram of a vehicle feature recognition system according to an embodiment of the present invention, and as shown in fig. 9, the system includes an enhancement module 62, a processing module 72, a false detection module 74, a screening module 76, a vehicle cover detection module 92, a license plate detection module 94, and a determination module 66, where the vehicle cover detection module 92 includes an extraction unit 82, a dimension reduction unit 84, a fusion unit 86, and a determination unit 88, and the license plate detection module 94 includes a recognition unit 96 and a determination unit 98.
When the vehicle feature recognition system performs image recognition, the method for vehicle feature recognition is completed, and in the case of performing nighttime muck vehicle feature recognition, fig. 10 is a flowchart three of the method for vehicle feature recognition according to the embodiment of the invention, and the method includes the following steps:
step S1002, acquiring a first identification image from a camera in a bayonet scene or an electric alarm scene, and sending the first identification image to an enhancement module to obtain a second identification image;
step S1004, inputting the second identification image into a processing module, wherein the processing module adopts a YOLO V3 algorithm, and sends the detected vehicle with low confidence coefficient into a false detection module to carry out secondary judgment on whether the vehicle is false detected or not, and directly eliminates the vehicle if the vehicle is false detected;
step S1006, the processed second identification image is sent to a vehicle cover detection module, HOG characteristics and CNN characteristics of the second identification image are extracted, a PCA algorithm is used for feature dimension reduction, and finally a linear regression model is used for predicting whether a cover exists or not;
step S1008, sending the processed second recognition image into a license plate detection module, detecting all characters capable of being detected, including Chinese characters, numbers and English, and finally judging whether a license plate exists in the module or not by combining confidence coefficient and the number of the detected characters;
and step S1010, under the conditions that the result of the vehicle cover detection module is strict and the result of the license plate detection module exists, marking the second identification image as safe, and performing linkage alarm on the second identification image with the unclean vehicle cover or the unclean license plate by amplifying.
The brightness of the first recognition image is improved through the enhancement module, the contrast of the first recognition image is enhanced, the second recognition image is obtained, the license plate recognition of the second recognition image is carried out, the judgment of warning is made according to the recognition result, the image enhancement module enables the image acquired at night to be processed and recognized more easily, the problem that the recognition degree of the muck vehicle features is low under the condition that the light at night is insufficient is solved, and the efficiency and the accuracy of the recognition of the image features of the vehicle at night are improved.
In one embodiment, a computer device is provided. The computer device may be a server. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used for storing image data. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a method of vehicle feature recognition.
In one embodiment, a computer device is provided, which may be a terminal. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a method of vehicle feature recognition. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the computer equipment, an external keyboard, a touch pad or a mouse and the like.
In one embodiment, a computer device is provided, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor executes the computer program to implement the steps of the method for identifying vehicle features provided in the above embodiments.
In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method for vehicle feature identification provided by the respective embodiments described above.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above examples only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A method of vehicle feature identification, the method comprising:
acquiring a first identification image of a vehicle, and processing preset parameters of the first identification image through an image enhancement model to obtain a second identification image;
acquiring geographic position information of a vehicle through a position recognition model, recognizing the second recognition image through a cover detection model, judging whether the cover of the vehicle in the second recognition image is covered tightly, recognizing the second recognition image through a license plate detection model, and judging whether the vehicle in the second recognition image has an amplified license plate, wherein the amplified license plate has complete and identifiable license plate information including the amplified license plate;
and under the condition that the hood detection result of the second identification image is not strict or the license plate detection result is absent, marking the second identification image as unsafe, uploading the geographic position information of the vehicle and sending out a warning signal.
2. The method of vehicle feature recognition according to claim 1, wherein before the recognizing the second recognition image by the hood detection model, the method further comprises:
segmenting and screening the second identification image through a target detection network algorithm to obtain a vehicle image in the second identification image and an image confidence coefficient of the vehicle image, and discarding the vehicle image corresponding to the image confidence coefficient under the condition that the image confidence coefficient is smaller than a first preset confidence coefficient;
under the condition that the image confidence coefficient is smaller than a second preset confidence coefficient and is larger than or equal to the first preset confidence coefficient, carrying out false detection analysis on the vehicle image through a convolutional neural network and a classifier algorithm, and judging whether the vehicle image is the image of the target vehicle or not according to the judgment result of the classifier algorithm;
and if the vehicle image is not the target vehicle, discarding the vehicle image corresponding to the image confidence.
3. The method of vehicle feature recognition according to claim 1, wherein the recognizing the second recognition image by the hood detection model includes:
extracting a first gradient feature and a first convolution neural network feature of the second identification image;
reducing the dimension of the first gradient feature to obtain a second gradient feature, and reducing the dimension of the first convolution neural network feature to obtain a second convolution neural network feature;
fusing the second gradient feature and the second convolutional neural network feature to obtain a fused feature;
and processing the fusion features through a linear regression model to obtain a fitting coefficient of the linear regression model, judging that the detection result of the vehicle cover is tight under the condition that the fitting coefficient is greater than or equal to a preset coefficient value, and judging that the detection result of the vehicle cover is not tight under the condition that the fitting coefficient is less than the preset coefficient value.
4. The method of vehicle feature recognition according to claim 1, wherein the recognizing the second recognition image by the license plate detection model comprises:
acquiring characters in the second recognition image to obtain the number of the characters and the confidence coefficient of the characters;
and judging that the license plate detection result exists when the number of the characters is greater than or equal to a preset number and the character confidence of each character is greater than or equal to a third preset confidence, and judging that the license plate detection result does not exist when the number of the characters is less than the preset number or the character confidence is less than the third preset confidence.
5. The method of vehicle feature recognition according to claim 1, wherein the obtaining vehicle geographic location information through a location recognition model comprises:
and acquiring the geographical position information of the monitoring equipment which captures the first identification image as the geographical position information of the vehicle.
6. The method for vehicle feature recognition according to claim 1, wherein the processing the preset parameters of the first recognition image through an image enhancement model to obtain a second recognition image comprises:
and carrying out contrast and brightness processing on the first identification image through an image enhancement model to obtain a second identification image, wherein the contrast and brightness processing comprises enhancing the contrast of the first identification image and improving the brightness of the first identification image.
7. The method of vehicle feature recognition according to claim 1, wherein after the second recognition image is recognized by a hood detection model, the method further comprises:
and under the conditions that the hood detection result of the second identification image is strict and the license plate detection result exists, marking the second identification image as safe.
8. An apparatus for vehicle feature recognition, the apparatus comprising an enhancement module, a detection module, and a determination module:
the enhancement module is used for acquiring a first identification image of a vehicle and processing the preset parameters of the first identification image through an image enhancement model to obtain a second identification image;
the detection module is used for acquiring vehicle geographic position information through a position recognition model, recognizing the second recognition image through a cover detection model, judging whether the cover of the vehicle in the second recognition image is tightly covered or not, recognizing the second recognition image through a license plate detection model, and judging whether an amplified license plate exists in the vehicle in the second recognition image or not, wherein the existence of the amplified license plate comprises the complete and identifiable license plate information of the amplified license plate;
the judging module is configured to mark the second recognition image as unsafe, upload the geographic position information of the vehicle, and issue a warning signal when the hood detection result of the second recognition image is not strict or the license plate detection result of the second recognition image is absent.
9. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911050662.9A CN110929589B (en) | 2019-10-31 | 2019-10-31 | Method, apparatus, computer apparatus and storage medium for identifying vehicle characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911050662.9A CN110929589B (en) | 2019-10-31 | 2019-10-31 | Method, apparatus, computer apparatus and storage medium for identifying vehicle characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110929589A true CN110929589A (en) | 2020-03-27 |
| CN110929589B CN110929589B (en) | 2023-07-07 |
Family
ID=69849973
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911050662.9A Active CN110929589B (en) | 2019-10-31 | 2019-10-31 | Method, apparatus, computer apparatus and storage medium for identifying vehicle characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110929589B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111860166A (en) * | 2020-06-18 | 2020-10-30 | 浙江大华技术股份有限公司 | Image detection method and device, computer equipment and storage medium |
| CN111881321A (en) * | 2020-07-27 | 2020-11-03 | 广元量知汇科技有限公司 | Smart city safety monitoring method based on artificial intelligence |
| CN112257541A (en) * | 2020-10-16 | 2021-01-22 | 浙江大华技术股份有限公司 | License plate recognition method, electronic device and computer-readable storage medium |
| CN112712708A (en) * | 2020-12-28 | 2021-04-27 | 上海眼控科技股份有限公司 | Information detection method, device, equipment and storage medium |
| CN113762012A (en) * | 2020-11-27 | 2021-12-07 | 北京京东尚科信息技术有限公司 | Illegal vehicle identification method and device, electronic equipment and medium |
| CN113762233A (en) * | 2020-06-05 | 2021-12-07 | 北京都是科技有限公司 | License plate detection method, system and device and thermal infrared image processor |
| CN113989722A (en) * | 2021-11-03 | 2022-01-28 | 北京赛博星通科技有限公司 | Muck truck compliance judgment method and system and storage medium |
| CN114078107A (en) * | 2020-08-11 | 2022-02-22 | 华晨宝马汽车有限公司 | Method, storage medium, apparatus, and system for identifying trim strips for vehicles |
| CN114115359A (en) * | 2021-10-26 | 2022-03-01 | 南京邮电大学 | A UAV mountain sheep hunting system and its working method |
| CN114648752A (en) * | 2020-12-21 | 2022-06-21 | 丰图科技(深圳)有限公司 | Accident vehicle license plate number detection method, device, electronic device and storage medium |
| CN114710970A (en) * | 2020-10-29 | 2022-07-05 | 华为技术有限公司 | Apparatus and method for locating a device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106845478A (en) * | 2016-12-30 | 2017-06-13 | 同观科技（深圳）有限公司 | The secondary licence plate recognition method and device of a kind of character confidence level |
| CN107784315A (en) * | 2016-08-26 | 2018-03-09 | 深圳光启合众科技有限公司 | The recognition methods of destination object and device, and robot |
| CN109409337A (en) * | 2018-11-30 | 2019-03-01 | 公安部交通管理科学研究所 | Muck vehicle feature identification method based on convolutional neural network |
| CN109508725A (en) * | 2017-09-15 | 2019-03-22 | 杭州海康威视数字技术股份有限公司 | Cover plate opening-closing detection method, device and the terminal of haulage vehicle |
| CN109815856A (en) * | 2019-01-08 | 2019-05-28 | 深圳中兴网信科技有限公司 | Status indication method, system and the computer readable storage medium of target vehicle |
-
2019
- 2019-10-31 CN CN201911050662.9A patent/CN110929589B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784315A (en) * | 2016-08-26 | 2018-03-09 | 深圳光启合众科技有限公司 | The recognition methods of destination object and device, and robot |
| CN106845478A (en) * | 2016-12-30 | 2017-06-13 | 同观科技（深圳）有限公司 | The secondary licence plate recognition method and device of a kind of character confidence level |
| CN109508725A (en) * | 2017-09-15 | 2019-03-22 | 杭州海康威视数字技术股份有限公司 | Cover plate opening-closing detection method, device and the terminal of haulage vehicle |
| CN109409337A (en) * | 2018-11-30 | 2019-03-01 | 公安部交通管理科学研究所 | Muck vehicle feature identification method based on convolutional neural network |
| CN109815856A (en) * | 2019-01-08 | 2019-05-28 | 深圳中兴网信科技有限公司 | Status indication method, system and the computer readable storage medium of target vehicle |
Non-Patent Citations (3)
| Title |
|---|
| 刘颖 等: "基于迁移学习及特征融合的轮胎花纹图像分类", 《计算机工程与设计》 * |
| 孙华魁 等: "《数字图像处理与识别技术研究》", 31 May 2019, 天津科学技术出版社 * |
| 柏森 等: "《信息隐藏算法及应用》", 30 September 2015, 国防工业出版社 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113762233A (en) * | 2020-06-05 | 2021-12-07 | 北京都是科技有限公司 | License plate detection method, system and device and thermal infrared image processor |
| CN111860166A (en) * | 2020-06-18 | 2020-10-30 | 浙江大华技术股份有限公司 | Image detection method and device, computer equipment and storage medium |
| CN111881321B (en) * | 2020-07-27 | 2021-04-20 | 东来智慧交通科技(深圳)有限公司 | A smart city security monitoring method based on artificial intelligence |
| CN111881321A (en) * | 2020-07-27 | 2020-11-03 | 广元量知汇科技有限公司 | Smart city safety monitoring method based on artificial intelligence |
| CN114078107A (en) * | 2020-08-11 | 2022-02-22 | 华晨宝马汽车有限公司 | Method, storage medium, apparatus, and system for identifying trim strips for vehicles |
| CN112257541A (en) * | 2020-10-16 | 2021-01-22 | 浙江大华技术股份有限公司 | License plate recognition method, electronic device and computer-readable storage medium |
| CN114710970A (en) * | 2020-10-29 | 2022-07-05 | 华为技术有限公司 | Apparatus and method for locating a device |
| CN113762012A (en) * | 2020-11-27 | 2021-12-07 | 北京京东尚科信息技术有限公司 | Illegal vehicle identification method and device, electronic equipment and medium |
| CN114648752A (en) * | 2020-12-21 | 2022-06-21 | 丰图科技(深圳)有限公司 | Accident vehicle license plate number detection method, device, electronic device and storage medium |
| CN112712708A (en) * | 2020-12-28 | 2021-04-27 | 上海眼控科技股份有限公司 | Information detection method, device, equipment and storage medium |
| CN114115359A (en) * | 2021-10-26 | 2022-03-01 | 南京邮电大学 | A UAV mountain sheep hunting system and its working method |
| CN114115359B (en) * | 2021-10-26 | 2024-06-14 | 南京邮电大学 | Unmanned aerial vehicle mountain area sheep searching system and working method thereof |
| CN113989722A (en) * | 2021-11-03 | 2022-01-28 | 北京赛博星通科技有限公司 | Muck truck compliance judgment method and system and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110929589B (en) | 2023-07-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110929589B (en) | Method, apparatus, computer apparatus and storage medium for identifying vehicle characteristics | |
| CN111667011B (en) | Damage detection model training and vehicle damage detection method, device, equipment and medium | |
| US9082038B2 (en) | Dram c adjustment of automatic license plate recognition processing based on vehicle class information | |
| CN106600977B (en) | Multi-feature recognition-based illegal parking detection method and system | |
| CN104616021B (en) | Traffic sign image processing method and device | |
| CN111507989A (en) | Training generation method of semantic segmentation model, and vehicle appearance detection method and device | |
| CN110706261A (en) | Vehicle violation detection method and device, computer equipment and storage medium | |
| US9460367B2 (en) | Method and system for automating an image rejection process | |
| CN109655075B (en) | Unmanned vehicle positioning method and device | |
| CN111444798B (en) | Identification method and device for driving behavior of electric bicycle and computer equipment | |
| CN112836683B (en) | License plate recognition method, device, equipment and medium for portable camera equipment | |
| CN113688805B (en) | Unmanned aerial vehicle-based unlicensed muck vehicle identification method and system | |
| CN113160272B (en) | Target tracking method and device, electronic equipment and storage medium | |
| CN111539317A (en) | Vehicle illegal driving detection method and device, computer equipment and storage medium | |
| CN111401282A (en) | Target detection method, target detection device, computer equipment and storage medium | |
| CN112016514B (en) | Traffic sign recognition method, device, equipment and storage medium | |
| CN111091041A (en) | Vehicle law violation judging method and device, computer equipment and storage medium | |
| CN111724408A (en) | Validation experiment method of abnormal driving behavior algorithm model based on 5G communication | |
| CN111753592A (en) | Traffic sign recognition method, device, computer equipment and storage medium | |
| KR101066081B1 (en) | In-vehicle smart information reading system and method | |
| Kiew et al. | Vehicle route tracking system based on vehicle registration number recognition using template matching algorithm | |
| CN111401362A (en) | Tampering detection method, device, equipment and storage medium for vehicle VIN code | |
| CN111476245A (en) | Vehicle left-turn violation detection method and device, computer equipment and storage medium | |
| CN113807125B (en) | Emergency lane occupation detection method, device, computer equipment and storage medium | |
| CN111462480B (en) | Traffic image evidence verification method and device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |