CN110738080A - method, device and electronic equipment for identifying modified motor vehicle - Google Patents
method, device and electronic equipment for identifying modified motor vehicle Download PDFInfo
- Publication number
- CN110738080A CN110738080A CN201810798044.1A CN201810798044A CN110738080A CN 110738080 A CN110738080 A CN 110738080A CN 201810798044 A CN201810798044 A CN 201810798044A CN 110738080 A CN110738080 A CN 110738080A
- Authority
- CN
- China
- Prior art keywords
- motor vehicle
- target area
- image
- neural network
- vehicle
- 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
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- 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 method for identifying the modified motor vehicle comprises the steps of locating a target region where the motor vehicle is located in an acquired image to be identified containing the motor vehicle, inputting the target region into a trained neural network, carrying out feature extraction and classification on the target region by the neural network, outputting the probability that the motor vehicle in the target region belongs to each preset motor vehicle attribute category, determining the preset motor vehicle attribute category corresponding to the maximum probability as the motor vehicle attribute category to which the motor vehicle in the target region belongs, and judging whether the motor vehicle in the image to be identified is the modified motor vehicle according to the output result of the neural network.
Description
Technical Field
The application relates to the technical field of image recognition, in particular to methods and devices for recognizing a modified motor vehicle and electronic equipment.
Background
However, for users, the original factory motor vehicle has a defect of , for example, the appearance or the performance of the motor vehicle cannot meet the requirements of the users.
Therefore, it is urgently needed to provide methods for identifying modified motor vehicles to assist in supervising the modification behaviors of the motor vehicles and guarantee the safety of travel.
Disclosure of Invention
In view of the above, the present application provides methods, apparatuses, and electronic devices for identifying a modified vehicle for screening the modified vehicle.
The present application provides, in an aspect, a method of identifying a refitted vehicle, comprising:
positioning a target area where the motor vehicle is located in the acquired image to be identified containing the motor vehicle;
inputting the target area into a trained th neural network, performing feature extraction and classification on the target area by the th neural network, outputting the probability that the motor vehicles in the target area belong to each preset motor vehicle attribute class, and determining the preset motor vehicle attribute class corresponding to the maximum probability as the motor vehicle attribute class to which the motor vehicles in the target area belong;
and judging whether the motor vehicle in the image to be identified is a motor vehicle to be refitted according to the output result of the th neural network.
A second aspect of the present application provides apparatus for identifying a modified vehicle, the apparatus comprising a location module, an identification module, and a processing module, wherein,
the positioning module is used for positioning a target area where the motor vehicle is located in the acquired image to be identified containing the motor vehicle;
the recognition module is used for inputting the target area into a trained th neural network, outputting the probability that the motor vehicles in the target area belong to each preset motor vehicle attribute category after performing feature extraction and classification on the target area by the th neural network, and determining the preset motor vehicle attribute category corresponding to the maximum probability as the motor vehicle attribute category to which the motor vehicles in the target area belong;
and the processing module is used for judging whether the motor vehicle in the image to be identified is a motor vehicle to be refitted according to the output result of the th neural network.
The third aspect of the present application provides computer readable storage media having stored thereon a computer program which when executed by a processor implements the steps of the method of identifying a refitted vehicle as claimed in any of the present application, aspect .
A fourth aspect of the present application provides electronic devices comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor when executing the program implementing the steps of the method of any of the present application for identifying a refitted vehicle as provided in the aspect.
According to the method, the device and the electronic equipment for identifying the refitted motor vehicle, the target area where the motor vehicle is located in the obtained image to be identified containing the motor vehicle, the target area is input into a trained -th neural network, the neural network is used for carrying out feature extraction and classification on the target area and then outputting the probability that the motor vehicle in the target area belongs to each preset motor vehicle attribute category, the preset motor vehicle attribute category corresponding to the maximum probability is determined as the motor vehicle attribute category to which the motor vehicle in the target area belongs, and then whether the motor vehicle in the image to be identified is the refitted motor vehicle or not is judged according to the output result of the neural network.
Drawings
Fig. 1A is a schematic illustration of a retrofit automotive vehicle based on appearance retrofitting as shown in an exemplary embodiment of the present application ;
FIG. 1B is a flow chart of an embodiment of a method for identifying a modified vehicle provided herein;
FIG. 2 is a flow chart of a second embodiment of a method for identifying a refitted vehicle provided herein;
FIG. 3 is a schematic diagram of an electronic device for identifying a device for modifying a vehicle according to an exemplary embodiment of the present application ;
FIG. 4 is a schematic structural diagram of an embodiment of an apparatus for identifying a modified vehicle provided herein;
fig. 5 is a schematic structural diagram of a second embodiment of the device for identifying a modified vehicle provided by the present application.
Detailed Description
The embodiments described in the exemplary embodiments below do not represent all embodiments consistent with the present application's patent, but rather are merely examples of apparatus and methods consistent with the present application's aspects patent, as detailed in the appended claims.
As used in this application and the appended claims, the singular forms "," "said," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It is to be understood that although the terms , second, third, etc. may be used herein to describe various information, these information should not be limited to these terms.
The application provides methods, devices and electronic equipment for identifying a refitted vehicle so as to screen the refitted vehicle.
The method, the device and the electronic equipment for identifying the refitted motor vehicle can be applied to a traffic security monitoring scene, so that manual intervention is reduced, labor cost is reduced, working efficiency is improved, motor vehicle refitting behaviors are assisted to be supervised, and travel safety is guaranteed.
Referring to fig. 1A, the appearance modification manner of the modified motor vehicle in the present application may include, but is not limited to, color modification, poster modification, skylight modification, adding a rear bar, adding a luggage rack, adding a rear tail wing, and the like.
The technical solution of the present application will be described in detail below with specific examples. The following several specific embodiments may be combined with each other, and details of the same or similar concepts or processes may not be repeated in some embodiments.
Referring to fig. 1B, a flow chart of an embodiment of a method for identifying a modified vehicle according to the present application is shown, where the method for identifying a modified vehicle according to the present embodiment includes:
s101, locating a target area where the motor vehicle is located in the acquired image to be identified containing the motor vehicle.
Specifically, a multi-target detection algorithm may be employed to locate a target region in which the vehicle is located in the image to be identified. For example, a target region where the vehicle is located in the image to be recognized may be located in the image to be recognized by using a DPM (Deformable Parts Model, DPM for short), an SSD (SSD) algorithm, a YOLO (young Only Look One, young joo) algorithm, a fast RCNN network, and the like. It should be noted that, for the specific implementation principle and implementation process of each algorithm, reference may be made to the description in the related art, and details are not described here.
S102, inputting the target area into a trained th neural network, performing feature extraction and classification on the target area by the th neural network, outputting the probability that the motor vehicle in the target area belongs to each preset motor vehicle attribute class, and determining the preset motor vehicle attribute class corresponding to the maximum probability as the motor vehicle attribute class to which the motor vehicle in the target area belongs, wherein the preset motor vehicle attribute classes at least comprise modified motor vehicles.
Specifically, for example, in embodiment, the th neural network includes two preset vehicle attribute categories, which are respectively a modified vehicle and another vehicle (where another category may be a non-vehicle or a pedestrian), in this step, after the target area is input into the th neural network, the th neural network performs feature extraction and classification on the target area, outputs a probability that the vehicle in the target area belongs to the modified vehicle and a probability that the vehicle in the target area belongs to another vehicle, and determines the preset vehicle attribute category corresponding to the maximum probability as the vehicle attribute category to which the vehicle in the target area belongs.
For another example, in another embodiment, the th neural network includes two preset vehicle attribute categories, the two preset vehicle attribute categories are respectively an refitted vehicle and a non-refitted vehicle, in this step, after the target region is input into the th neural network, the th neural network performs feature extraction and classification on the target region, outputs the probability that the vehicle in the target region belongs to the refitted vehicle and the probability that the vehicle in the target region belongs to the non-refitted vehicle, and determines the preset vehicle attribute category corresponding to the maximum probability as the vehicle attribute category to which the vehicle in the target region belongs.
It should be noted that the training process of the neural network will be described in detail in the following embodiments, and will not be described herein.
S103, judging whether the motor vehicle in the image to be identified is a converted motor vehicle or not according to the output result of the th neural network.
Specifically, in this step, when the output result of the neural network indicates that the vehicle attribute category to which the vehicle in the target area belongs is the converted vehicle, the vehicle in the image to be recognized is determined to be the converted vehicle.
The method for identifying the refitted motor vehicle provided by the embodiment includes the steps of locating a target area where the motor vehicle is located in an obtained image to be identified containing the motor vehicle, inputting the target area into a trained -th neural network, performing feature extraction and classification on the target area through the -th neural network, outputting probabilities that the motor vehicle in the target area belongs to each preset motor vehicle attribute category, determining the preset motor vehicle attribute category corresponding to the maximum probability as the motor vehicle attribute category to which the motor vehicle in the target area belongs, and judging whether the motor vehicle in the image to be identified is the refitted motor vehicle according to an output result of the -th neural network.
Fig. 2 is a flowchart of a second embodiment of the method for identifying a refitted vehicle provided by the present application. Referring to fig. 2, the method provided in this embodiment may include:
s201, at least kinds of processing in white balance processing and image enhancement processing are carried out on the acquired image to be identified containing the motor vehicle.
For example, in an embodiment, when the white balance processing is performed on the image to be recognized, the pixel value of the pixel point whose pixel value exceeds 220 may be updated to 220 when the average pixel value of the image to be recognized is greater than 220, and the pixel value of each pixel point may be updated to a current value plus a specified value when the average pixel value of the image to be recognized is less than 100.
S202, positioning a target area where the motor vehicle is located in the image to be identified.
Specifically, the specific implementation process and implementation principle of this step may refer to the description in the foregoing embodiments, and are not described herein again.
S203, inputting the target area into a trained th neural network, performing feature extraction and classification on the target area by the th neural network, outputting the probability that the motor vehicle in the target area belongs to each preset motor vehicle attribute class, and determining the preset motor vehicle attribute class corresponding to the maximum probability as the motor vehicle attribute class to which the motor vehicle in the target area belongs, wherein the preset motor vehicle attribute classes at least comprise modified motor vehicles.
For example, in embodiment, the th neural network includes three preset vehicle attribute categories, which are respectively a modified vehicle, a non-modified vehicle and others, in this step, after the target area is input into the th neural network, the th neural network performs feature extraction and classification on the target area and outputs a probability that the vehicle in the target area belongs to the modified vehicle, a probability that the vehicle in the target area belongs to the non-modified vehicle and a probability that the vehicle in the target area belongs to the others, and determines the preset vehicle attribute category corresponding to the maximum probability as the vehicle attribute category to which the vehicle in the target area belongs.
And S204, detecting whether the image corresponding to the target area is abnormal or not.
Specifically, whether the image corresponding to the target area is abnormal or not has an important influence on identifying whether the motor vehicle is a converted motor vehicle or not. Therefore, in this embodiment, a step of detecting whether the image corresponding to the target area is abnormal is added. In specific implementation, a traditional method can be adopted to detect whether the image corresponding to the target area is abnormal, and when the image corresponding to the target area is detected to be distorted, inclined, fuzzy, seriously shielded, excessively exploded, excessively dark or seriously lost, the image corresponding to the target area is determined to be abnormal.
Optionally, in a possible implementation manner of the present application , a specific implementation process of the step may include:
(1) inputting the target region into a trained second neural network, performing feature extraction and classification on the target region by the second neural network, outputting the probability that the target region belongs to each preset image quality category, and determining the preset image quality category corresponding to the maximum probability as the image quality category to which the target region belongs; the preset image quality categories include normal and abnormal.
(2) And determining the output result of the second neural network as a detection result for detecting whether the image corresponding to the imaging of the target area is abnormal or not.
For example, in the embodiment, after the target region is input into the second neural network, the second neural network performs feature extraction and classification on the target region, and then the probability that the output target region belongs to an abnormality is 0.7, and the probability that the output target region belongs to a normal state is 0.3, at this time, it is determined that the image quality class to which the target region belongs is an abnormality, that is, the output result of the second neural network is that the image quality class to which the target region belongs is an abnormality, and at this time, it is determined that the detection result of detecting whether the image corresponding to the target region is an abnormality.
In this embodiment, the execution sequence of step S203 and step S204 is not limited, and step S203 and step S204 may be executed simultaneously.
S205, determining whether the motor vehicle in the image to be identified is the motor vehicle to be refitted according to the output result of the th neural network and the detection result of detecting whether the image corresponding to the target area is abnormal.
Specifically, when the -th neural network outputs that the attribute class of the motor vehicle belonging to the motor vehicle in the target area is the converted motor vehicle and the detection result is that the image corresponding to the target area is normal, the motor vehicle in the image to be identified is determined to be the converted motor vehicle.
The method provided by the embodiment determines that the motor vehicle in the image to be identified is the converted motor vehicle when the detection result is that the image corresponding to the target area is normal and the output result of the neural network is that the motor vehicle attribute category to which the motor vehicle in the target area belongs is the converted motor vehicle.
, when the neural network includes 3 preset categories (the 3 preset categories are modified motor vehicles, non-modified motor vehicles and others (wherein, the other categories can be non-motor vehicles or pedestrians and the like), at this time, when the output result of the neural network is that the motor vehicle attribute categories to which the motor vehicles in the target area belong are other, determining that the category to which the motor vehicles in the image to be identified belong is unknown;
when the output result of the neural network indicates that the attribute class of the motor vehicle belonging to the motor vehicle in the target area is the non-refitted motor vehicle, determining that the motor vehicle in the image to be identified is the non-refitted motor vehicle;
and when the output result of the neural network indicates that the attribute class of the motor vehicle belonging to the motor vehicle in the target area is the modified motor vehicle and the detection result indicates that the image corresponding to the target area is abnormal, determining that the motor vehicle in the image to be identified is the non-modified motor vehicle.
That is, when the output result of the th neural network indicates that the attribute class of the motor vehicle to which the motor vehicle in the target area belongs is other, it is determined that the class to which the motor vehicle in the image to be recognized belongs is unknown regardless of the detection result of whether the image corresponding to the detection target area is abnormal or not.
And when the output result of the neural network indicates that the motor vehicle attribute category to which the motor vehicle in the target area belongs is the non-refitted motor vehicle, determining that the motor vehicle in the image to be identified is the non-refitted motor vehicle no matter what the detection result of whether the image corresponding to the target area imaging is abnormal or not.
When the image corresponding to the target area where the motor vehicle is located is abnormal, the image cannot normally reflect the actual appearance of the motor vehicle, and therefore the motor vehicle cannot be identified on the basis of the image.
According to the method for identifying the refitted motor vehicle, whether the motor vehicle in the image to be identified is the refitted motor vehicle or not is determined according to the output result of the th neural network and the detection result of whether the image corresponding to the detection target area is abnormal or not by detecting whether the image corresponding to the target area is abnormal or not.
It should be noted that th neural network and the second neural network in the present application are pre-trained neural networks.
(1) Building neural networks
Specifically, in the present application, the th neural network and the second neural network are mainly used for feature extraction and classification, and the network results are similar to the structure of the neural network used for feature extraction and classification in the related art, and are not described herein again.
(2) Obtaining training samples
For example, for the th neural network, a sample image containing a motor vehicle can be acquired (i.e. a positive sample image containing a normal motor vehicle and a negative sample image containing a modified motor vehicle are acquired), and then the acquired sample image is labeled and labeled.
For another example, for the second neural network, an image including a normal vehicle may be acquired, and the acquired image may be labeled and tagged.
(3) Training the neural network by using the obtained training sample to obtain the trained neural network
Specifically, the network parameters in the neural network may be set to specified values, and then the obtained training samples are used to train the neural network, so as to obtain the trained neural network.
Specifically, the process can comprise two stages of forward propagation and backward propagation, wherein the forward propagation is to input training samples, forward propagation is performed on the training samples to extract data characteristics, a loss function is calculated, backward propagation is performed, the loss function is sequentially performed from the last layer of the neural network forward and backward, and meanwhile, the network parameters of the neural network are modified by a gradient descent method, so that the loss function is converged.
In accordance with the foregoing embodiments of the method of identifying a refitted vehicle, the present application also provides embodiments of an apparatus for identifying a refitted vehicle.
In terms of hardware, as shown in fig. 3, for the hardware structure diagram of the electronic device where the apparatus for identifying the modified motor vehicle shown in the exemplary embodiment of the application is located, except for the memory 310, the processor 320, the memory 330, and the network interface 340 shown in fig. 3, the electronic device where the apparatus is located generally modifies actual functions of the apparatus for modifying the motor vehicle according to the identification, and may further include other hardware, which is not described again.
Fig. 4 is a schematic structural diagram of an embodiment of the apparatus for identifying a modified vehicle provided by the present application, referring to fig. 4, the apparatus provided by the present application may include a positioning module 410, an identification module 420, and a processing module 430, wherein,
the positioning module 410 is configured to position a target area where the motor vehicle is located in the acquired image to be identified, which includes the motor vehicle;
the identification module 420 is configured to input the target region into a trained th neural network, output probabilities that motor vehicles in the target region belong to each preset motor vehicle attribute category after performing feature extraction and classification on the target region by the th neural network, and determine a preset motor vehicle attribute category corresponding to a maximum probability as a motor vehicle attribute category to which the motor vehicles in the target region belong;
and the processing module 430 is configured to determine whether the vehicle in the image to be identified is a modified vehicle according to the output result of the th neural network.
The apparatus of this embodiment may be used to implement the technical solution of the method embodiment shown in fig. 1B, and the implementation principle and the technical effect are similar, which are not described herein again.
, fig. 5 is a schematic structural diagram of a second embodiment of the apparatus for identifying a modified vehicle provided by the present application, referring to fig. 5, the apparatus provided by the present embodiment may further include a detection module 440 based on the above embodiment, wherein,
the detection module 440 is configured to detect whether an image corresponding to a target area is abnormal after the positioning module 410 positions the target area where the motor vehicle is located in the acquired image to be identified containing the motor vehicle;
the processing module 430 is specifically configured to determine whether the motor vehicle in the image to be identified is a motor vehicle to be reinstalled according to the output result of the -th neural network and the detection result of detecting whether the image corresponding to the target area is abnormal.
, the processing module 430 is specifically configured to determine that the vehicle in the image to be identified is the reinstalled vehicle when the output result of the th neural network indicates that the vehicle attribute category to which the vehicle in the target area belongs is the reinstalled vehicle, and the detection result indicates that the image corresponding to the target area is normal.
, the processing module 430 is further configured to:
when the output result of the neural network indicates that the attribute class of the motor vehicle in the target area is a non-refitted motor vehicle, determining that the motor vehicle in the image to be identified is the non-refitted motor vehicle;
and when the output result of the th neural network indicates that the attribute class of the motor vehicle belonging to the motor vehicle in the target area is the motor vehicle to be converted, and the detection result indicates that the image corresponding to the target area is abnormal, determining that the motor vehicle in the image to be identified is the motor vehicle to be converted.
, the detecting module 440 is specifically configured to input the target region into a trained second neural network, perform feature extraction and classification on the target region by the second neural network, output probabilities that the target region belongs to each preset image quality category, determine a preset image quality category corresponding to the maximum probability as the image quality category to which the target region belongs, and determine an output result of the second neural network as a detection result of detecting whether an image corresponding to the target region is abnormal, where the preset image quality categories include normal and abnormal.
, the processing module 430 is further configured to perform at least of white balance processing and enhancement processing on the image to be recognized before the positioning module 410 locates the target area where the motor vehicle is located in the acquired image to be recognized containing the motor vehicle.
The present application further provides computer readable storage media having stored thereon a computer program that, when executed by a processor, performs the steps of any of the methods for identifying a retrofit vehicle provided herein .
In particular, computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media and memory devices, including by way of example semiconductor memory devices (e.g., EPROM, EEPROM, and flash memory devices), magnetic disks (e.g., internal hard disk or removable disks), magneto-optical disks, and CD ROM and DVD-ROM disks.
Continuing to refer to fig. 3, the present application further provides electronic devices comprising a memory 310, a processor 320, and a computer program stored on the memory 310 and executable on the processor 320, the processor implementing the steps of the method for identifying a refitted vehicle as provided herein at any .
The above description is only exemplary of the present application and should not be taken as limiting the present application, as any modification, equivalent replacement, or improvement made within the spirit and principle of the present application should be included in the scope of protection of the present application.
Claims (14)
1, A method of identifying a refitted vehicle, the method comprising:
positioning a target area where the motor vehicle is located in the acquired image to be identified containing the motor vehicle;
inputting the target area into a trained th neural network, performing feature extraction and classification on the target area by the th neural network, outputting the probability that the motor vehicles in the target area belong to each preset motor vehicle attribute class, and determining the preset motor vehicle attribute class corresponding to the maximum probability as the motor vehicle attribute class to which the motor vehicles in the target area belong;
and judging whether the motor vehicle in the image to be identified is a motor vehicle to be refitted according to the output result of the th neural network.
2. The method according to claim 1, wherein after locating the target area in which the motor vehicle is located in the acquired image to be identified containing the motor vehicle, the method further comprises:
detecting whether the image corresponding to the target area is abnormal or not;
the judging whether the motor vehicle in the image to be identified is a motor vehicle to be refitted according to the output result of the th neural network comprises the following steps:
and determining whether the motor vehicle in the image to be identified is a motor vehicle to be refitted according to the output result of the th neural network and the detection result of detecting whether the image corresponding to the target area is abnormal.
3. The method according to claim 2, wherein the determining whether the motor vehicle in the image to be identified is a refitted motor vehicle according to the output result of the neural network and the detection result of whether the image corresponding to the target area is abnormal comprises:
and when the output result of the th neural network indicates that the attribute class of the motor vehicle belonging to the motor vehicle in the target area is the motor vehicle to be converted, and the detection result indicates that the image corresponding to the target area is normal, determining that the motor vehicle in the image to be identified is the motor vehicle to be converted.
4. The method according to claim 3, wherein the determining whether the vehicle in the image to be identified is a modified vehicle according to the output result of the neural network and the detection result of detecting whether the image corresponding to the target area is abnormal further includes:
when the output result of the neural network indicates that the attribute class of the motor vehicle in the target area is a non-refitted motor vehicle, determining that the motor vehicle in the image to be identified is the non-refitted motor vehicle;
and when the output result of the th neural network indicates that the attribute class of the motor vehicle belonging to the motor vehicle in the target area is the motor vehicle to be converted, and the detection result indicates that the image corresponding to the target area is abnormal, determining that the motor vehicle in the image to be identified is the motor vehicle to be converted.
5. The method according to claim 2, wherein the detecting whether the image corresponding to the target area is abnormal includes:
inputting the target region into a trained second neural network, performing feature extraction and classification on the target region by the second neural network, outputting the probability that the target region belongs to each preset image quality category, and determining the preset image quality category corresponding to the maximum probability as the image quality category to which the target region belongs; the preset image quality categories comprise normal and abnormal;
and determining the output result of the second neural network as a detection result for detecting whether the image corresponding to the target area is abnormal or not.
6. The method according to claim 1, wherein before locating the target area in which the motor vehicle is located in the acquired image to be identified containing the motor vehicle, the method further comprises:
and performing at least of white balance processing and enhancement processing on the image to be recognized.
An apparatus for identifying a refitted vehicle of the kind 7, , comprising a location module, an identification module and a processing module, wherein,
the positioning module is used for positioning a target area where the motor vehicle is located in the acquired image to be identified containing the motor vehicle;
the recognition module is used for inputting the target area into a trained th neural network, outputting the probability that the motor vehicles in the target area belong to each preset motor vehicle attribute class after performing feature extraction and classification on the target area by the th neural network, and determining the preset motor vehicle attribute class corresponding to the maximum probability as the motor vehicle attribute class to which the motor vehicles in the target area belong;
and the processing module is used for judging whether the motor vehicle in the image to be identified is a motor vehicle to be refitted according to the output result of the th neural network.
8. The device according to claim 7, further comprising a detection module, wherein the detection module is configured to detect whether the image corresponding to the target area is abnormal after the positioning module positions the target area where the motor vehicle is located in the acquired image to be identified containing the motor vehicle;
the processing module is specifically configured to determine whether the motor vehicle in the image to be identified is a motor vehicle to be reinstalled according to the output result of the -th neural network and the detection result of detecting whether the image corresponding to the target area is abnormal.
9. The device according to claim 8, wherein the processing module is specifically configured to determine that the vehicle in the image to be identified is a remoter vehicle when the output result of the th neural network is that the vehicle attribute category to which the vehicle in the target area belongs is a remoter vehicle and the detection result is that the image corresponding to the target area is normal.
10. The apparatus of claim 9, wherein the processing module is further configured to:
when the output result of the neural network indicates that the attribute class of the motor vehicle in the target area is a non-refitted motor vehicle, determining that the motor vehicle in the image to be identified is the non-refitted motor vehicle;
and when the output result of the th neural network indicates that the attribute class of the motor vehicle belonging to the motor vehicle in the target area is the motor vehicle to be converted, and the detection result indicates that the image corresponding to the target area is abnormal, determining that the motor vehicle in the image to be identified is the motor vehicle to be converted.
11. The apparatus according to claim 8, wherein the detection module is specifically configured to input the target region into a trained second neural network, perform feature extraction and classification on the target region by the second neural network, output probabilities that the target region belongs to each preset image quality category, and determine a preset image quality category corresponding to a maximum probability as the image quality category to which the target region belongs; determining the output result of the second neural network as the detection result for detecting whether the image corresponding to the target area is abnormal or not; wherein the preset image quality categories include normal and abnormal.
12. The device according to claim 7, wherein the processing module is further configured to perform at least of white balance processing and enhancement processing on the image to be recognized before the positioning module locates the target area where the motor vehicle is located in the acquired image to be recognized containing the motor vehicle.
A computer readable storage medium , having a computer program stored thereon, wherein the program, when executed by a processor, performs the steps of the method of any of claims 1-6 to .
14, an electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method of any of claims 1-6 when executing the program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810798044.1A CN110738080A (en) | 2018-07-19 | 2018-07-19 | method, device and electronic equipment for identifying modified motor vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810798044.1A CN110738080A (en) | 2018-07-19 | 2018-07-19 | method, device and electronic equipment for identifying modified motor vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110738080A true CN110738080A (en) | 2020-01-31 |
Family
ID=69235644
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810798044.1A Pending CN110738080A (en) | 2018-07-19 | 2018-07-19 | method, device and electronic equipment for identifying modified motor vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110738080A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111291812A (en) * | 2020-02-11 | 2020-06-16 | 浙江大华技术股份有限公司 | Attribute class acquisition method and device, storage medium and electronic device |
| CN111310837A (en) * | 2020-02-21 | 2020-06-19 | 广州华工邦元信息技术有限公司 | Vehicle refitting recognition method, device, system, medium and equipment |
| CN111444813A (en) * | 2020-03-23 | 2020-07-24 | 北京字节跳动网络技术有限公司 | Method, device, equipment and storage medium for identifying attribute classification of target object |
| CN113348663A (en) * | 2021-03-23 | 2021-09-03 | 深圳市锐明技术股份有限公司 | Container monitoring method, terminal device and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105488453A (en) * | 2015-11-30 | 2016-04-13 | 杭州全实鹰科技有限公司 | Detection identification method of no-seat-belt-fastening behavior of driver based on image processing |
| CN106295601A (en) * | 2016-08-18 | 2017-01-04 | 合肥工业大学 | A kind of Safe belt detection method of improvement |
| CN107578034A (en) * | 2017-09-29 | 2018-01-12 | 百度在线网络技术（北京）有限公司 | information generating method and device |
| CN107730898A (en) * | 2016-11-08 | 2018-02-23 | 北京奥斯达兴业科技有限公司 | Parking lot illegal vehicle recognition methods and system |
| CN107958200A (en) * | 2017-10-13 | 2018-04-24 | 上海眼控科技股份有限公司 | A kind of intelligent checking system and method for car repacking exhaust fan |
| CN107967446A (en) * | 2017-10-13 | 2018-04-27 | 上海眼控科技股份有限公司 | A kind of intelligent checking system and method for installing engine protection device additional |
| CN107967445A (en) * | 2017-10-13 | 2018-04-27 | 上海眼控科技股份有限公司 | A kind of car installs the intelligent checking system and method for skylight additional |
-
2018
- 2018-07-19 CN CN201810798044.1A patent/CN110738080A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105488453A (en) * | 2015-11-30 | 2016-04-13 | 杭州全实鹰科技有限公司 | Detection identification method of no-seat-belt-fastening behavior of driver based on image processing |
| CN106295601A (en) * | 2016-08-18 | 2017-01-04 | 合肥工业大学 | A kind of Safe belt detection method of improvement |
| CN107730898A (en) * | 2016-11-08 | 2018-02-23 | 北京奥斯达兴业科技有限公司 | Parking lot illegal vehicle recognition methods and system |
| CN107578034A (en) * | 2017-09-29 | 2018-01-12 | 百度在线网络技术（北京）有限公司 | information generating method and device |
| CN107958200A (en) * | 2017-10-13 | 2018-04-24 | 上海眼控科技股份有限公司 | A kind of intelligent checking system and method for car repacking exhaust fan |
| CN107967446A (en) * | 2017-10-13 | 2018-04-27 | 上海眼控科技股份有限公司 | A kind of intelligent checking system and method for installing engine protection device additional |
| CN107967445A (en) * | 2017-10-13 | 2018-04-27 | 上海眼控科技股份有限公司 | A kind of car installs the intelligent checking system and method for skylight additional |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111291812A (en) * | 2020-02-11 | 2020-06-16 | 浙江大华技术股份有限公司 | Attribute class acquisition method and device, storage medium and electronic device |
| CN111291812B (en) * | 2020-02-11 | 2023-10-17 | 浙江大华技术股份有限公司 | Method and device for acquiring attribute category, storage medium and electronic device |
| CN111310837A (en) * | 2020-02-21 | 2020-06-19 | 广州华工邦元信息技术有限公司 | Vehicle refitting recognition method, device, system, medium and equipment |
| CN111444813A (en) * | 2020-03-23 | 2020-07-24 | 北京字节跳动网络技术有限公司 | Method, device, equipment and storage medium for identifying attribute classification of target object |
| CN113348663A (en) * | 2021-03-23 | 2021-09-03 | 深圳市锐明技术股份有限公司 | Container monitoring method, terminal device and storage medium |
| CN113348663B (en) * | 2021-03-23 | 2023-01-24 | 深圳市锐明技术股份有限公司 | Container monitoring method, terminal device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10474713B1 (en) | Learning method and learning device using multiple labeled databases with different label sets and testing method and testing device using the same | |
| CN110738080A (en) | method, device and electronic equipment for identifying modified motor vehicle | |
| CN110020592B (en) | Object detection model training method, device, computer equipment and storage medium | |
| EP3796228A1 (en) | Device and method for generating a counterfactual data sample for a neural network | |
| JP5127392B2 (en) | Classification boundary determination method and classification boundary determination apparatus | |
| CN110706261A (en) | Vehicle violation detection method and device, computer equipment and storage medium | |
| CN110909692A (en) | Abnormal license plate recognition method and device, computer storage medium and electronic equipment | |
| US11487295B2 (en) | Method and device for abstracting a data record | |
| CN110874471B (en) | Privacy and safety protection neural network model training method and device | |
| CN111626367A (en) | Countermeasure sample detection method, apparatus, device and computer readable storage medium | |
| CN111539317A (en) | Vehicle illegal driving detection method and device, computer equipment and storage medium | |
| CN109543647A (en) | A kind of road abnormality recognition method, device, equipment and medium | |
| CN115100739A (en) | Man-machine behavior detection method, system, terminal device and storage medium | |
| CN111476102A (en) | Safety protection method, central control equipment and computer storage medium | |
| CN113887306A (en) | Driving scene perception detection method and device, electronic equipment and storage medium | |
| US20230350664A1 (en) | Methods and apparatus for automatically labeling data processing events in autonomous driving vehicles via machine learning | |
| US11403498B2 (en) | Method for classifying a capture taken by a sensor | |
| CN113011454A (en) | System and method for robust active learning method using noisy labels and domain adaptation | |
| CN111652137A (en) | Illegal vehicle detection method and device, computer equipment and storage medium | |
| CN112182269A (en) | Training of image classification model, image classification method, device, equipment and medium | |
| CN114419584A (en) | Improved traffic sign identification and positioning method by inhibiting YOLOv4 by using non-maximum value | |
| CN112396126B (en) | Target detection method and system based on detection trunk and local feature optimization | |
| CN114359669A (en) | Picture analysis model adjusting method and device and computer readable storage medium | |
| CN111340261B (en) | Method, system, computer equipment and storage medium for judging order violation | |
| CN113723431A (en) | Image recognition method, image recognition device and computer-readable 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 |