CN111814751A - Vehicle attribute analysis method and system based on deep learning target detection and image recognition - Google Patents
Vehicle attribute analysis method and system based on deep learning target detection and image recognition Download PDFInfo
- Publication number
- CN111814751A CN111814751A CN202010817195.4A CN202010817195A CN111814751A CN 111814751 A CN111814751 A CN 111814751A CN 202010817195 A CN202010817195 A CN 202010817195A CN 111814751 A CN111814751 A CN 111814751A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- image
- attribute analysis
- deep learning
- analysis method
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
-
- 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
- G06V20/59—Context or environment of the image inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
-
- 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
Abstract
The invention discloses a vehicle attribute analysis method and system based on deep learning target detection and image recognition, which comprises the following steps: a video acquisition step, which is used for acquiring video image information of the bayonet and acquiring and uploading video acquisition equipment; a vehicle detection step for extracting vehicle image information in the image acquired in the video acquisition step, wherein the vehicle image information refers to position coordinates of a vehicle in the image; a database writing step; and in the step of searching the image by the image, extracting the characteristics of the vehicle from the vehicle attribute analysis module, wherein the characteristics refer to the characteristics obtained from a specific certain layer in the deep learning network, so as to represent the vehicle, and the characteristics are used for retrieving similar vehicles according to the provided vehicle images and providing specific information stored in a database of the similar vehicles. The invention adopts the technology of searching the images by the images, thereby improving the database retrieval speed; aiming at the passing specificity of the vehicles at the bayonet, only the information of the vehicle at the front is extracted from each image, so that the model precision is ensured while the model speed is improved.
Description
Technical Field
The invention relates to a computer vision, pattern recognition, video processing, artificial intelligence and intelligent monitoring system, in particular to a vehicle attribute analysis system based on deep learning target detection and image recognition.
Background
In recent years, with the rapid development of economy and society, the living standard of people is continuously improved, the requirements on the living quality are higher and higher, and the happiness index is continuously improved. As an important component of people's daily lives, the appearance of tools directly affects the happiness index. With the wide spread of automobiles and the increase of urban vehicles, urban traffic is under increasing pressure. The problems of traffic jam, traffic accidents and the like are more serious while the convenience is brought to daily travel of people. Various vehicle violations are frequent, particularly casualties caused by traffic accidents are high in successive years, but a large number of traffic accidents exist.
Taking the tracking of hit-and-run vehicles as an example, a common method is to take surrounding monitoring information according to the occurrence position of a car accident as a starting point, manually screen the hit-and-run direction of the hit-and-run vehicles, and gradually track the vehicles according to a monitoring video. However, the manual screening speed is slow, the position of an escaping vehicle is difficult to accurately determine, the passing vehicles cannot be screened automatically, particularly, the vehicles need to be manually checked one by one when the vehicles are out of a gate, traffic jam is easily caused, and a vicious circle is generated. The traditional method is time-consuming and labor-consuming, needs a large amount of resources and personnel allocation, and is not enough in execution force. In particular, the continuous development of artificial intelligence and deep learning, whether hardware resources or software resources, has satisfied the prerequisite condition of falling on the ground of artificial intelligence products, and therefore, it is necessary to provide a vehicle attribute analysis system based on deep learning target detection and image recognition. The vehicle screening method is not only beneficial to vehicle screening, but also can play a positive role in subsequent extension applications.
In the prior art, Zhongming adopts a plurality of models to respectively detect and identify the attributes of pedestrians and vehicles in the invention patent application 'a video image structured analysis system facing the field of intelligent security protection', the deployment of an actual scene is difficult under the condition of first resources, and the cost is relatively high; in the invention patent application of Chenchang Jian, namely 'vehicle structural information extraction based on deep learning', only vehicle attributes and driver related information are extracted, but only a license plate on the surface of a vehicle body can be used as a unique identifier of the vehicle. In the invention patent application of the Yanfan, namely 'vehicle license plate automatic detection method and system based on vehicle lamp identification in dark environment', vehicle information is obtained through vehicle lamp characteristics, generally, in dark environment, the vehicle lamp is relatively bright, the effect is seriously influenced, meanwhile, corresponding vehicle registration information needs to be called according to the vehicle license plate information, and the application range of the invention is limited.
In the prior art, the methods have certain defects. The invention overcomes the defects and shortcomings in the invention, adopts multi-label classification on the selection of the depth model, uses the depth learning model as little as possible, and simultaneously extracts only one maximum target vehicle on each picture according to the specificity of the bayonet, thereby greatly improving the speed and the precision of the model. In the vehicle search, the map searching technology is adopted, a plurality of target vehicles can be searched in a database according to similarity sequence according to the provided vehicle information, and the elapsed time is given. Specifically, the invention relates to a vehicle attribute analysis system based on deep learning target detection and image recognition, which comprises: the video acquisition module is used for acquiring video image information of the bayonet; the vehicle detection module extracts the vehicle image collected in the video collection module; the vehicle attribute analysis module is used for analyzing the vehicle attributes detected by the vehicle detection module, wherein the vehicle attributes comprise vehicle colors, vehicle brands, vehicle sub-brands and vehicle types; the vehicle feature detection module is used for detecting a main driving safety belt and a secondary driving safety belt of a vehicle picture, a sun shield, an annual inspection label, a vehicle pendant and a paper towel box decoration of the vehicle picture obtained by the vehicle detection module; and the license plate recognition module is used for detecting the license plate of the vehicle image detected by the vehicle detection module and giving the license plate number and the license plate type.
Disclosure of Invention
1. Objects of the invention
The invention aims to provide a vehicle attribute analysis system based on deep learning target detection and image recognition aiming at the fields of intelligent security, monitoring and the like, and aims to automatically extract attribute information of a vehicle and store related information into a database for a map searching module to read by utilizing a deep learning technology.
2. The technical scheme adopted by the invention
The invention discloses a vehicle attribute analysis method based on deep learning target detection and image recognition, which comprises the following steps of:
a video acquisition step, which is used for acquiring video image information of the bayonet and acquiring and uploading video acquisition equipment;
a vehicle detection step for extracting vehicle image information in the image acquired in the video acquisition step, wherein the vehicle image information refers to position coordinates of a vehicle in the image;
a database writing step, which is used for writing the vehicle structural information and the corresponding vehicle images obtained by the vehicle detection module, the vehicle attribute analysis module, the license plate recognition module and the vehicle feature detection module into a database for being read by the image searching module;
in the step of searching the map, the features of the vehicle are extracted from the vehicle attribute analysis module, the features refer to features obtained from one layer in a deep learning network, and for a classification network, the features are usually obtained from a full connection layer, so that the vehicle is represented, the vehicle is used for retrieving similar vehicles according to the provided vehicle images, and specific information stored in a database by the similar vehicles is provided, specifically:
establishing an image data set and an image feature library;
extracting features of the image to be retrieved through a feature extraction network to obtain features with the retrieved image;
and calculating the similarity of the image to be retrieved to each feature in the feature library under a similarity measurement criterion, and finally sequencing the images according to the similarity and sequentially outputting corresponding pictures.
Preferably, in the vehicle detection step, when the vehicle passes through the gate, only the front vehicle is extracted from a single image and determined according to the vehicle coordinates.
Preferably, only the front vehicle is extracted from a single image, the vehicle position frame is expanded according to a certain proportion, and complete vehicle image information is cut out.
Preferably, the complete vehicle image information includes vehicle attributes, vehicle features, license plate identification.
Preferably, the vehicle attribute comprises vehicle body color identification, vehicle brand identification, vehicle sub-brand identification and vehicle type identification;
preferably, the vehicle features comprise a primary and secondary driving safety belt, a primary and secondary driving sun shield, a vehicle annual inspection label, a vehicle interior hanging ornament and a tissue box ornament.
Preferably, the license plate identification comprises a license plate number and a license plate type, wherein the license plate type is that the license plates are classified according to colors.
Preferably, the vehicle detection step determines a maximum target position of the image passing through the gate, and determines the maximum target position based on coordinate information of the vehicle.
Preferably, the license plate recognition step is through Chinese character, digit, English and color recognition.
The invention discloses a vehicle attribute analysis system based on deep learning target detection and image recognition, which comprises a memory and a processor and is used for the vehicle attribute analysis method.
3. Advantageous effects adopted by the present invention
(1) By adopting multi-label classification, a plurality of attribute information of the vehicle can be obtained at the same time, resource consumption and repeated calculation when a plurality of models are adopted are reduced, and the calculation speed is effectively improved;
(2) the vehicle attribute analysis system comprises license plate identification representing a unique identifier of a vehicle, and can accurately position information of a specific vehicle according to license plate information;
(3) the database retrieval speed is improved by adopting the image searching technology;
(4) aiming at the passing specificity of the vehicles at the bayonet, only the information of the vehicle at the front is extracted from each image, so that the model precision is ensured while the model speed is improved.
Drawings
FIG. 1 is a schematic diagram of an overall architecture of a vehicle attribute analysis system based on deep learning target detection and image recognition;
FIG. 2 is a schematic diagram of a vehicle attribute analysis module in a vehicle attribute analysis system based on deep learning target detection and image recognition;
FIG. 3 is a schematic diagram of a vehicle feature detection module in a vehicle attribute analysis system based on deep learning target detection and image recognition;
FIG. 4 is a schematic diagram of a map searching module in a vehicle attribute analysis system based on deep learning target detection and image recognition;
FIG. 5 is a first diagram of vehicle attribute analysis effects in a vehicle attribute analysis system based on deep learning target detection and image recognition;
FIG. 6 is a second diagram of the vehicle attribute analysis effect in a vehicle attribute analysis system based on deep learning target detection and image recognition;
fig. 7 is a third diagram of the effect of vehicle attribute analysis in a vehicle attribute analysis system based on deep learning target detection and image recognition.
Detailed Description
The technical solutions in the examples of the present invention are clearly and completely described below with reference to the drawings in the examples of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without inventive step, are within the scope of the present invention.
The present invention will be described in further detail with reference to the accompanying drawings.
Example 1
As shown in fig. 1, the overall architecture of the vehicle attribute analysis system based on deep learning target detection and image recognition mainly includes several modules, which are respectively: the system comprises a video acquisition module, a vehicle detection module, a vehicle attribute analysis module, a vehicle feature detection module, a license plate recognition module, a database writing module and a picture searching module. The video acquisition module is used for acquiring video image information of the bayonet and taking the video image information as the input of the car face detection module; the vehicle detection module is used for detecting the position coordinates of the vehicle in the image information acquired by the video acquisition module, and externally expanding the target frame according to a certain proportion so as to cut out a complete vehicle image and serve as the input of a subsequent vehicle attribute analysis module, a vehicle feature object detection module and a license plate recognition module; the vehicle attribute analysis module is used for analyzing the attributes of the vehicle image detected by the vehicle detection module, and specifically comprises vehicle color, vehicle brand, vehicle sub-brand and vehicle type; the vehicle feature detection module is used for detecting the position coordinates of a main driving safety belt, a secondary driving safety belt, a sun shield, an annual inspection label, a vehicle pendant hanging ornament and a tissue box ornament in a vehicle image obtained by the vehicle detection module; the license plate recognition module is used for detecting the license plate of the vehicle image detected by the vehicle detection module and giving a license plate number and a license plate type, wherein the license plate type refers to the ground color of the license plate; the database writing module is used for writing vehicle structural information such as vehicle characteristics and the like obtained by the vehicle attribute analysis module, the license plate recognition module, the vehicle characteristic object detection module and a specific layer in the vehicle attribute analysis network into a database for being read by the image searching module; and the image searching module is used for searching the vehicle information written in the database according to the similarity indexes according to the provided vehicle images.
Specifically, taking the toll gate as an example, it is obvious that the description is only a part of embodiments of the present invention, but not all embodiments, and the detailed steps are as follows:
and 2, vehicle detection, namely extracting vehicle image information in the image acquired by the video acquisition module according to the toll gate video acquired by the video acquisition module, wherein the vehicle image information refers to the position coordinates of the vehicle in the image. Normally, a vehicle can only pass one vehicle at a time when passing through a toll gate, so for the accuracy of the information, only the front-most vehicle is needed on a single image, and the front-most vehicle is generally determined according to the coordinates of the vehicle. Meanwhile, in order to ensure that the vehicle coordinate frame contains all vehicle information, the vehicle position frame needs to be expanded according to a certain proportion so as to cut out complete vehicle image information;
step 4, detecting a vehicle feature object detection module shown in fig. 3, detecting a primary and secondary driving safety belt, a primary and secondary driving sun shield, a vehicle annual inspection label, a vehicle interior hanging decoration, a tissue box decoration and the like on a vehicle picture according to the vehicle image information detected and cut by the vehicle detection module;
step 5, license plate recognition, namely recognizing license plate numbers and license plate types on the vehicle images obtained by cutting according to the vehicle image information detected and cut by the vehicle detection module, wherein the license plate types refer to license plates classified according to colors;
step 6, writing in a database, writing the vehicle structural information and the corresponding vehicle images obtained by the vehicle detection module, the vehicle attribute analysis module, the license plate recognition module and the vehicle feature detection module into the database for the image searching module to read, wherein the information specifically comprises: the vehicle detection module detects the position coordinates of the vehicle, the vehicle image cut after the vehicle image is expanded according to the proportion, the color of the vehicle body, the vehicle brand, the vehicle sub-brand, the vehicle type, the license plate number, the license plate type, the passing time and the like;
and 7, searching the map, and extracting the features of the vehicle from the vehicle attribute analysis module according to the related information obtained by the module, wherein the features refer to the features obtained by a specific layer in the deep learning network, and the layer is not limited to a specific mode, such as the last full connection layer in a ResNet34 network structure, the last full connection layer in a VGG16 network structure and the like. And searching similar vehicles according to the similarity ranking through the provided vehicle images, and giving specific information stored in the database by the similar vehicles. The flow of searching the diagram is shown in fig. 4, and the specific steps are as follows:
step 7-1: establishing an image data set and an image feature library, extracting image features by using a convolutional neural network, such as ResNet34 in the embodiment, and outputting the last full connection layer in the network
Step 7-2: uploading an image to be retrieved, and extracting features through a feature extraction network to obtain features with a retrieved image;
and 7-3: and calculating the similarity of the image to be retrieved to each feature in the feature library under a certain similarity measurement criterion, and finally sequencing the images according to the similarity and sequentially outputting corresponding pictures. In this embodiment, the euclidean distance is selected as the similarity measurement criterion, and if the euclidean distance between the image to be retrieved and the image in the database is smaller, the similarity between the two images is higher.
As shown in fig. 5-7, by adopting the multi-tag classification, the license plate identification including the unique identifier representing the vehicle, and the map searching technology, a plurality of attribute information of the vehicle can be obtained simultaneously, resource consumption and repeated calculation when a plurality of models are adopted are reduced, the database retrieval speed is increased, the calculation speed is effectively increased, only the foremost vehicle information is extracted from each image according to the passing specificity of vehicles at the checkpoint, and the model accuracy is ensured while the model speed is increased.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A vehicle attribute analysis method based on deep learning target detection and image recognition is characterized by comprising the following steps:
a video acquisition step, which is used for acquiring video image information of the bayonet and acquiring and uploading video acquisition equipment;
a vehicle detection step for extracting vehicle image information in the image acquired in the video acquisition step, wherein the vehicle image information refers to position coordinates of a vehicle in the image;
a database writing step, which is used for writing the vehicle structural information and the corresponding vehicle images obtained by the vehicle detection module, the vehicle attribute analysis module, the license plate recognition module and the vehicle feature detection module into a database for being read by the image searching module;
in the step of searching the map, the features of the vehicle are extracted from the vehicle attribute analysis module, the features refer to features obtained from one layer in a deep learning network, and for a classification network, the features are usually obtained from a full connection layer, so that the vehicle is represented, the vehicle is used for retrieving similar vehicles according to the provided vehicle images, and specific information stored in a database by the similar vehicles is provided, specifically:
establishing an image data set and an image feature library;
extracting features of the image to be retrieved through a feature extraction network to obtain features with the retrieved image;
and calculating the similarity of the image to be retrieved to each feature in the feature library under a similarity measurement criterion, and finally sequencing the images according to the similarity and sequentially outputting corresponding pictures.
2. The deep learning target detection and image recognition-based vehicle attribute analysis method according to claim 1, characterized in that: and a vehicle detection step, wherein when the vehicle passes through the gate, only the front vehicle is extracted from the single image and is determined according to the vehicle coordinates.
3. The deep learning target detection and image recognition-based vehicle attribute analysis method according to claim 2, characterized in that: and only the front vehicle is extracted from a single image, the vehicle position frame is expanded according to a certain proportion, and complete vehicle image information is cut out.
4. The method of claim 3, wherein the complete vehicle image information comprises vehicle attributes, vehicle features, and license plate recognition.
5. The deep learning target detection and image recognition-based vehicle attribute analysis method according to claim 4, characterized in that: the vehicle attributes include vehicle body color identification, vehicle brand identification, vehicle sub-brand identification, vehicle type identification.
6. The deep learning target detection and image recognition-based vehicle attribute analysis method according to claim 4, characterized in that: the vehicle characteristic objects comprise a main driving safety belt, a main driving sunshade plate, a vehicle annual inspection label, a vehicle inner hanging piece hanging decoration and a tissue box decoration.
7. The deep learning target detection and image recognition-based vehicle attribute analysis method according to claim 4, characterized in that: the license plate identification comprises a license plate number and a license plate type, wherein the license plate type is that license plates are classified according to colors.
8. The deep learning target detection and image recognition-based vehicle attribute analysis method according to claim 1, characterized in that: and the vehicle detection step is to determine the maximum target position of the image passing through the gate and determine the maximum target position according to the coordinate information of the vehicle.
9. The vehicle attribute analysis method based on deep learning target detection and image recognition as claimed in claim 1, wherein the license plate recognition step is through chinese character, digit, english, color recognition.
10. A vehicle attribute analysis system based on deep learning object detection and image recognition, comprising a memory and a processor, using the vehicle attribute analysis method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010817195.4A CN111814751A (en) | 2020-08-14 | 2020-08-14 | Vehicle attribute analysis method and system based on deep learning target detection and image recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010817195.4A CN111814751A (en) | 2020-08-14 | 2020-08-14 | Vehicle attribute analysis method and system based on deep learning target detection and image recognition |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111814751A true CN111814751A (en) | 2020-10-23 |
Family
ID=72858983
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010817195.4A Pending CN111814751A (en) | 2020-08-14 | 2020-08-14 | Vehicle attribute analysis method and system based on deep learning target detection and image recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111814751A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113190703A (en) * | 2021-04-02 | 2021-07-30 | 深圳市安软科技股份有限公司 | Intelligent retrieval method and device for video image, electronic equipment and storage medium |
| CN113269150A (en) * | 2021-06-24 | 2021-08-17 | 浪潮云信息技术股份公司 | Vehicle multi-attribute identification system and method based on deep learning |
| CN113360523A (en) * | 2020-12-10 | 2021-09-07 | 四川警察学院 | Multiple screening system based on vehicle big data |
| CN115050005A (en) * | 2022-06-17 | 2022-09-13 | 北京精英路通科技有限公司 | Target detection method and detection device for high-level video intelligent parking scene |
| WO2022262013A1 (en) * | 2021-06-15 | 2022-12-22 | 合肥维天运通信息科技股份有限公司 | Method and system for collecting and monitoring vehicle state information |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060030985A1 (en) * | 2003-10-24 | 2006-02-09 | Active Recognition Technologies Inc., | Vehicle recognition using multiple metrics |
| CN103530366A (en) * | 2013-10-12 | 2014-01-22 | 湖北微模式科技发展有限公司 | Vehicle searching method and system based on user-defined features |
| CN105808732A (en) * | 2016-03-10 | 2016-07-27 | 北京大学 | Integration target attribute identification and precise retrieval method based on depth measurement learning |
| CN106776943A (en) * | 2016-12-01 | 2017-05-31 | 中科唯实科技（北京）有限公司 | A kind of vehicle retrieval method based on AutoEncoder and attribute tags |
| CN107103314A (en) * | 2017-06-15 | 2017-08-29 | 浙江南自智能科技股份有限公司 | A kind of fake license plate vehicle retrieval system based on machine vision |
| CN107730905A (en) * | 2017-06-13 | 2018-02-23 | 银江股份有限公司 | Multitask fake license plate vehicle vision detection system and method based on depth convolutional neural networks |
| CN108108657A (en) * | 2017-11-16 | 2018-06-01 | 浙江工业大学 | A kind of amendment local sensitivity Hash vehicle retrieval method based on multitask deep learning |
| CN108595576A (en) * | 2018-04-17 | 2018-09-28 | 高新兴科技集团股份有限公司 | It is a kind of based on the vehicle of database to scheme to search drawing method |
| CN108984644A (en) * | 2018-06-25 | 2018-12-11 | 高新兴科技集团股份有限公司 | A kind of fake-licensed car search method and system |
| CN109325420A (en) * | 2018-08-27 | 2019-02-12 | 广州烽火众智数字技术有限公司 | A kind of vehicle characteristics search method and system based on big data |
| CN109508395A (en) * | 2018-09-26 | 2019-03-22 | 安徽四创电子股份有限公司 | A method of the intelligence based on vehicle retrieval is combined related cases |
| CN110400464A (en) * | 2019-07-15 | 2019-11-01 | 杭州亿圣信息技术有限公司 | A kind of cell disobeys the monitoring method of parking |
| CN110659374A (en) * | 2019-09-19 | 2020-01-07 | 江苏鸿信系统集成有限公司 | Method for searching images by images based on neural network extraction of vehicle characteristic values and attributes |
| CN110807415A (en) * | 2019-10-31 | 2020-02-18 | 南通大学 | Traffic checkpoint vehicle intelligent retrieval system and method based on annual inspection marks |
| CN110876090A (en) * | 2018-09-04 | 2020-03-10 | 杭州海康威视数字技术股份有限公司 | Video abstract playback method and device, electronic equipment and readable storage medium |
-
2020
- 2020-08-14 CN CN202010817195.4A patent/CN111814751A/en active Pending
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060030985A1 (en) * | 2003-10-24 | 2006-02-09 | Active Recognition Technologies Inc., | Vehicle recognition using multiple metrics |
| CN103530366A (en) * | 2013-10-12 | 2014-01-22 | 湖北微模式科技发展有限公司 | Vehicle searching method and system based on user-defined features |
| CN105808732A (en) * | 2016-03-10 | 2016-07-27 | 北京大学 | Integration target attribute identification and precise retrieval method based on depth measurement learning |
| CN106776943A (en) * | 2016-12-01 | 2017-05-31 | 中科唯实科技（北京）有限公司 | A kind of vehicle retrieval method based on AutoEncoder and attribute tags |
| CN107730905A (en) * | 2017-06-13 | 2018-02-23 | 银江股份有限公司 | Multitask fake license plate vehicle vision detection system and method based on depth convolutional neural networks |
| CN107103314A (en) * | 2017-06-15 | 2017-08-29 | 浙江南自智能科技股份有限公司 | A kind of fake license plate vehicle retrieval system based on machine vision |
| CN108108657A (en) * | 2017-11-16 | 2018-06-01 | 浙江工业大学 | A kind of amendment local sensitivity Hash vehicle retrieval method based on multitask deep learning |
| CN108595576A (en) * | 2018-04-17 | 2018-09-28 | 高新兴科技集团股份有限公司 | It is a kind of based on the vehicle of database to scheme to search drawing method |
| CN108984644A (en) * | 2018-06-25 | 2018-12-11 | 高新兴科技集团股份有限公司 | A kind of fake-licensed car search method and system |
| CN109325420A (en) * | 2018-08-27 | 2019-02-12 | 广州烽火众智数字技术有限公司 | A kind of vehicle characteristics search method and system based on big data |
| CN110876090A (en) * | 2018-09-04 | 2020-03-10 | 杭州海康威视数字技术股份有限公司 | Video abstract playback method and device, electronic equipment and readable storage medium |
| CN109508395A (en) * | 2018-09-26 | 2019-03-22 | 安徽四创电子股份有限公司 | A method of the intelligence based on vehicle retrieval is combined related cases |
| CN110400464A (en) * | 2019-07-15 | 2019-11-01 | 杭州亿圣信息技术有限公司 | A kind of cell disobeys the monitoring method of parking |
| CN110659374A (en) * | 2019-09-19 | 2020-01-07 | 江苏鸿信系统集成有限公司 | Method for searching images by images based on neural network extraction of vehicle characteristic values and attributes |
| CN110807415A (en) * | 2019-10-31 | 2020-02-18 | 南通大学 | Traffic checkpoint vehicle intelligent retrieval system and method based on annual inspection marks |
Non-Patent Citations (2)
| Title |
|---|
| NA JIANG 等: "Multi-Attribute Driven Vehicle Re-Identification with Spatial-Temporal Re-Ranking", 《2018 25TH IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING》, pages 858 - 862 * |
| 何霞;汤一平;陈朋;王丽冉;袁公萍;: "多任务分段紧凑特征的车辆检索方法", 《中国图象图形学报》, vol. 23, no. 12, pages 1801 - 1812 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113360523A (en) * | 2020-12-10 | 2021-09-07 | 四川警察学院 | Multiple screening system based on vehicle big data |
| CN113190703A (en) * | 2021-04-02 | 2021-07-30 | 深圳市安软科技股份有限公司 | Intelligent retrieval method and device for video image, electronic equipment and storage medium |
| WO2022262013A1 (en) * | 2021-06-15 | 2022-12-22 | 合肥维天运通信息科技股份有限公司 | Method and system for collecting and monitoring vehicle state information |
| US11954955B2 (en) | 2021-06-15 | 2024-04-09 | Hefei Weitianyuntong Information Technology Co., Ltd | Method and system for collecting and monitoring vehicle status information |
| CN113269150A (en) * | 2021-06-24 | 2021-08-17 | 浪潮云信息技术股份公司 | Vehicle multi-attribute identification system and method based on deep learning |
| CN115050005A (en) * | 2022-06-17 | 2022-09-13 | 北京精英路通科技有限公司 | Target detection method and detection device for high-level video intelligent parking scene |
| CN115050005B (en) * | 2022-06-17 | 2024-04-05 | 北京精英路通科技有限公司 | Target detection method and detection device for high-level video intelligent parking scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111814751A (en) | Vehicle attribute analysis method and system based on deep learning target detection and image recognition | |
| CN107729818B (en) | Multi-feature fusion vehicle re-identification method based on deep learning | |
| CN109857889B (en) | Image retrieval method, device and equipment and readable storage medium | |
| CN108090429B (en) | Vehicle type recognition method for graded front face bayonet | |
| CN110378236B (en) | Vehicle identity recognition model construction and recognition method and system based on deep learning | |
| De Charette et al. | Real time visual traffic lights recognition based on spot light detection and adaptive traffic lights templates | |
| CN102289686B (en) | Method for identifying classes of moving targets based on transfer learning | |
| CN109871799B (en) | Method for detecting mobile phone playing behavior of driver based on deep learning | |
| CN105160309A (en) | Three-lane detection method based on image morphological segmentation and region growing | |
| CN113420607A (en) | Multi-scale target detection and identification method for unmanned aerial vehicle | |
| CN104615986A (en) | Method for utilizing multiple detectors to conduct pedestrian detection on video images of scene change | |
| CN103310006B (en) | A kind of area-of-interest exacting method in vehicle DAS (Driver Assistant System) | |
| CN108647695A (en) | Soft image conspicuousness detection method based on covariance convolutional neural networks | |
| CN110659374A (en) | Method for searching images by images based on neural network extraction of vehicle characteristic values and attributes | |
| CN113963399A (en) | Personnel trajectory retrieval method and device based on multi-algorithm fusion application | |
| CN114818819A (en) | Road obstacle detection method based on millimeter wave radar and visual signal | |
| Kiew et al. | Vehicle route tracking system based on vehicle registration number recognition using template matching algorithm | |
| CN109993049A (en) | A kind of video image structure analysis system towards intelligent security guard field | |
| CN111753610A (en) | Weather identification method and device | |
| CN104463238A (en) | License plate recognition method and system | |
| Coronado et al. | Detection and classification of road signs for automatic inventory systems using computer vision | |
| CN113360523A (en) | Multiple screening system based on vehicle big data | |
| Priambada et al. | Levensthein distance as a post-process to improve the performance of ocr in written road signs | |
| CN114120286A (en) | Traffic signal lamp identification method and device | |
| CN113283303A (en) | License plate recognition method and device |
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 |