CN109325488A - For assisting the method, device and equipment of car damage identification image taking - Google Patents
For assisting the method, device and equipment of car damage identification image taking Download PDFInfo
- Publication number
- CN109325488A CN109325488A CN201811013914.6A CN201811013914A CN109325488A CN 109325488 A CN109325488 A CN 109325488A CN 201811013914 A CN201811013914 A CN 201811013914A CN 109325488 A CN109325488 A CN 109325488A
- Authority
- CN
- China
- Prior art keywords
- information
- detection model
- component
- shooting
- image
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/08—Insurance
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Development Economics (AREA)
- Technology Law (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Multimedia (AREA)
- Studio Devices (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
This specification embodiment provides a kind of method for assisting car damage identification image taking, device and equipment, this specification embodiment, by the image for obtaining photographing module acquisition, identify the component of target vehicle in image, and the relative pose of component information the detection photographing module and target vehicle obtained according at least to identification, obtaining includes posture information one or more in shooting distance information and information of shooting angles, based on the comparison result that posture information obtained is compared to acquisition with preset expected pose information, output is for guiding user's control photographing module with the prompting message of expected pose photographic subjects vehicle, style of shooting can be adjusted with guiding photographing personnel.
Description
Technical field
This specification is related to technical field of data processing, more particularly, to the method for auxiliary car damage identification image taking,
Device and equipment.
Background technique
In vehicle insurance industry, when car accident proposition Claims Resolution application occurs for car owner, insurance company needs the damage journey to vehicle
Degree is assessed, to determine the project list for needing to repair and compensate the amount of money.Currently, carrying out setting loss process to the vehicle that is in danger
In, the foundation material of core is car damage identification image.
Car damage identification image is usually to carry out acquisition of taking pictures by operating personnel scene at present, then according to the photograph of live shooting
Piece carries out car damage identification processing.Car damage identification image is often required clearly reflect damaged vehicle situation, this is usually
It needs the personnel of taking pictures that there is the relevant knowledge of car damage identification, the image for obtaining and meeting setting loss processing requirement could be shot.However it is real
In the scene of border, often shot actively or in the case where insurance company operating personnel requires by car owner, the car damage identification image of acquisition
Setting loss image processing requirements may not met.The car damage identification image of different shooting qualities can obtain different setting loss results.
Summary of the invention
To overcome the problems in correlation technique, present description provides for assisting the side of car damage identification image taking
Method, device and equipment.
According to this specification embodiment in a first aspect, provide a kind of method for assisting car damage identification image taking,
The described method includes:
Obtain the image of photographing module acquisition；
Identify the component of target vehicle in described image, and the component information obtained according at least to identification detects photographing module
With the relative pose of target vehicle, posture information is obtained, the posture information includes shooting distance information and information of shooting angles
One of or it is a variety of；
Based on the comparison result that the posture information is compared to acquisition with preset expected pose information, output is used for
Guide user's control photographing module with the prompting message of expected pose photographic subjects vehicle.
In one embodiment, the component information include: component locations, component sizes and component mark, and/or, institute
Stating shooting distance information is photographing module affiliated distance range at a distance from target vehicle, and/or, the information of shooting angles is
The affiliated angular range of shooting angle.
In one embodiment, the component information is based on knowing described image using preset component detection model
It does not obtain；The component detection model is based on being trained initial part detection model using the first training sample data obtaining
?；In the first training sample data, sample characteristics include sample image, and sample label includes vehicle part in sample image
Component information.
In one embodiment, the shooting distance information is based on: with the output data of the component detection model and
Described image carries out prediction acquisition apart from detection model as the preset input data apart from detection model, and using described；
It is described to be based on being trained acquisition to initial distance detection model using the second training sample data apart from detection model；Second
In training sample data, sample characteristics include sample image, in sample image vehicle part component information, sample label includes
Shooting distance information.
In one embodiment, the information of shooting angles is based on: with output data, the distance of the component detection model
Input data of the output data and described image of detection model as preset angle detection model, and utilize the angle
Detection model carries out prediction acquisition；The angle detection model is based on detecting mould to initial angle using third training sample data
Type is trained acquisition；In third training sample data, sample characteristics include sample image, vehicle part in sample image
Component information and shooting distance information, sample label include information of shooting angles.
In one embodiment, the initial part detection model, the initial distance detection model and initial angle
Detection model respectively includes MobileNets model.
In one embodiment, described shot with expected pose includes: to carry out pan-shot, close shot with specified shooting angle
One of shooting, the shooting of middle scape etc. are a variety of, pan-shot, the shooting of middle scape and close shot shooting by shooting distance from big to small
Sequence is divided.
According to the second aspect of this specification embodiment, provide it is a kind of for assisting the device of car damage identification image taking,
Described device includes:
Image collection module is used for: obtaining the image of photographing module acquisition；
Information detecting module is used for: the component of target vehicle in identification described image, and according at least to the portion that identification obtains
The relative pose of part infomation detection photographing module and target vehicle obtains posture information, and the posture information includes shooting distance
One of information and information of shooting angles are a variety of；
Information reminding module, is used for: based on the posture information is compared acquisition with preset expected pose information
Comparison result, export for guiding user's control photographing module with the prompting message of expected pose photographic subjects vehicle.
In one embodiment, the component information include: component locations, component sizes and component mark, and/or, institute
Stating shooting distance information is photographing module affiliated distance range at a distance from target vehicle, and/or, the information of shooting angles is
The affiliated angular range of shooting angle.
In one embodiment, the component information is based on knowing described image using preset component detection model
It does not obtain；The component detection model is based on being trained initial part detection model using the first training sample data obtaining
?；In the first training sample data, sample characteristics include sample image, and sample label includes vehicle part in sample image
Component information.
In one embodiment, the shooting distance information is based on: with the output data of the component detection model and
Described image carries out prediction acquisition apart from detection model as the preset input data apart from detection model, and using described；
It is described to be based on being trained acquisition to initial distance detection model using the second training sample data apart from detection model；Second
In training sample data, sample characteristics include sample image, in sample image vehicle part component information, sample label includes
Shooting distance information.
In one embodiment, the information of shooting angles is based on: with output data, the distance of the component detection model
Input data of the output data and described image of detection model as preset angle detection model, and utilize the angle
Detection model carries out prediction acquisition；The angle detection model is based on detecting mould to initial angle using third training sample data
Type is trained acquisition；In third training sample data, sample characteristics include sample image, vehicle part in sample image
Component information and shooting distance information, sample label include information of shooting angles.
In one embodiment, the initial part detection model, the initial distance detection model and initial angle
Detection model respectively includes MobileNets model.
In one embodiment, described shot with expected pose includes: to carry out pan-shot, close shot with specified shooting angle
Shooting, middle scape shooting etc. one of or it is a variety of, the pan-shot, middle scape shooting and close shot shooting by shooting distance from greatly to
Small sequence is divided.
According to the third aspect of this specification embodiment, a kind of computer equipment is provided, including memory, processor and deposit
Store up the computer program that can be run on a memory and on a processor, wherein the processor is realized when executing described program
Following method:
Obtain the image of photographing module acquisition；
Identify the component of target vehicle in described image, and the component information obtained according at least to identification detects photographing module
With the relative pose of target vehicle, posture information is obtained, the posture information includes shooting distance information and information of shooting angles
One of or it is a variety of；
Based on the comparison result that the posture information is compared to acquisition with preset expected pose information, output is used for
Guide user's control photographing module with the prompting message of expected pose photographic subjects vehicle.
The technical solution that the embodiment of this specification provides can include the following benefits:
This specification embodiment, by obtaining the image of photographing module acquisition, the component of target vehicle in identification image, and
According at least to the relative pose of component information detection photographing module and target vehicle that identification obtains, obtaining includes that shooting distance is believed
One or more posture information in breath and information of shooting angles, is based on posture information obtained and preset expected pose
Information is compared the comparison result of acquisition, exports for guiding user's control photographing module with expected pose photographic subjects vehicle
Prompting message, thus realize the posture information of photographing module is fed back, can with guiding photographing personnel adjust style of shooting,
Improve the shooting quality of car damage identification image.
It should be understood that above general description and following detailed description be only it is exemplary and explanatory, not
This specification can be limited.
Detailed description of the invention
The drawings herein are incorporated into the specification and forms part of this specification, and shows the reality for meeting this specification
Example is applied, and is used to explain the principle of this specification together with specification.
Fig. 1 is a kind of this specification application scenarios for shooting car damage identification image shown according to an exemplary embodiment
Figure.
Fig. 2 is that this specification is shown according to an exemplary embodiment a kind of for assisting the side of car damage identification image taking
The flow chart of method.
Fig. 3 A is that this specification is shown according to an exemplary embodiment another for assisting car damage identification image taking
Method flow chart.
Fig. 3 B is that this specification is shown according to an exemplary embodiment a kind of for assisting car damage identification image taking
Application example.
Fig. 4 is that this specification is shown according to an exemplary embodiment a kind of for assisting the dress of car damage identification image taking
A kind of hardware structure diagram of computer equipment where setting.
Fig. 5 is that this specification is shown according to an exemplary embodiment a kind of for assisting the dress of car damage identification image taking
The block diagram set.
Specific embodiment
Example embodiments are described in detail here, and the example is illustrated in the accompanying drawings.Following description is related to
When attached drawing, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements.Following exemplary embodiment
Described in embodiment do not represent all embodiments consistent with this specification.On the contrary, they are only and such as institute
The example of the consistent device and method of some aspects be described in detail in attached claims, this specification.
It is only to be not intended to be limiting this explanation merely for for the purpose of describing particular embodiments in the term that this specification uses
Book.The "an" of used singular, " described " and "the" are also intended to packet in this specification and in the appended claims
Most forms are included, unless the context clearly indicates other meaning.It is also understood that term "and/or" used herein is
Refer to and includes that one or more associated any or all of project listed may combine.
It will be appreciated that though various information may be described using term first, second, third, etc. in this specification, but
These information should not necessarily be limited by these terms.These terms are only used to for same type of information being distinguished from each other out.For example, not taking off
In the case where this specification range, the first information can also be referred to as the second information, and similarly, the second information can also be claimed
For the first information.Depending on context, word as used in this " if " can be construed to " ... when " or
" when ... " or " in response to determination ".
Car insurance setting loss, can be by science, specialized inspection, test and the exploration means of system, to vehicle collision
Comprehensive analysis is carried out with the scene of the accident, with vehicle assessment of loss data and mantenance data, scientific system is carried out to vehicle collision reparation
The assessment of loss price.And car damage identification image is one of vehicle assessment of loss data, car damage identification image can be for the vehicle that is in danger
The image for carrying out setting loss core damage, carries out acquisition of taking pictures to scene often through operating personnel or car owner.In order to clearly react
The information such as the concrete position, defective component, type of impairment of damaged vehicle, degree of injury out, it is past to the quality of car damage identification image
It is strict with toward having.And with the rapid development of mobile terminal, user can utilize the mobile end with shooting function at any time
End is shot.As shown in Figure 1, being a kind of this specification shooting car damage identification image shown according to an exemplary embodiment
Application scenario diagram.User can use mobile phone and take pictures to damaged vehicle, obtain car damage identification image.However, by amateur
Personnel shoot damaged vehicle, and the car damage identification image of acquisition may not meet setting loss image processing requirements.And different bats
The car damage identification image for taking the photograph quality will directly affect final setting loss result, it is therefore desirable to which providing a kind of can improve car damage identification figure
As the processing scheme of shooting quality.
This specification embodiment provide it is a kind of for assisting the scheme of car damage identification image taking, pass through increase image taking
Guiding function can adjust style of shooting with guiding photographing personnel, improve the shooting quality of car damage identification image, be subsequent setting loss core
Damage Claims Resolution process provides accurate car damage identification image, to generate more accurate car damage identification result.
This specification embodiment is illustrated below in conjunction with attached drawing.
As shown in Fig. 2, being that this specification is shown according to an exemplary embodiment a kind of for assisting car damage identification image
The flow chart of the method for shooting, which comprises
In step 202, the image of photographing module acquisition is obtained；
In step 204, the component of target vehicle in described image is identified, and according at least to the component information that identification obtains
Detect photographing module and target vehicle relative pose, obtain posture information, the posture information include shooting distance information and
One of information of shooting angles is a variety of；
In step 206, based on by the posture information compared with preset expected pose information is compared acquisition
As a result, output is for guiding user's control photographing module with the prompting message of expected pose photographic subjects vehicle.
Wherein, acquired image can be stored image, for example, be triggered by control of taking pictures and the figure that saves
Picture；Acquired image is also possible to photographing module and currently acquires but also not stored image.In one embodiment, by default
Frequency carries out the image that photographing module acquires to cut frame processing, intercepts photographing module acquired image.For example, predeterminated frequency can
To be 1 second 2 times.In another embodiment, photographing module acquired image can also be obtained, in real time so as in real time to user
Current camera lens shooting picture is detected, and judges whether that user is needed to adjust shooting posture.
The purpose of this specification embodiment is to judge the opposite position of photographing module and target vehicle using the image currently obtained
The relativeness of appearance, i.e. shooting module and target vehicle.Relative pose can be shooting distance, be also possible to shooting angle, lead to
It crosses identified shooting distance or shooting angle, is compared with desired shooting distance or shooting angle, and according to comparing
As a result the prompting message of output guidance user's shooting carries out Real-time Feedback to the photographing information of photographing module to realize, is used for
Correct the improper problem for leading to picture quality difference of style of shooting.
In order to improve the accuracy rate of posture information, the present embodiment can first identify that the component in image on target vehicle is believed
Breath.Component can be the component part of composition target vehicle.Component information can be the information of description component, in one embodiment
In, component information may include component locations, component sizes and component mark.Component locations can be the position of component in image
It sets.Component sizes can be the size of component in image.Component mark can be for distinguishing different components on target vehicle
Mark, for example, component mark can be name of parts coding.In one embodiment, component sizes and component locations can pass through
Component rectangle frame is labeled, and indicates component sizes and component locations using component rectangle frame location coordinate information.In other realities
Apply in example, component information can also including component shape etc. information, will not repeat them here.
It is illustrated by taking a kind of component information recognition methods as an example below.
The component information is based on carrying out identification acquisition to described image using preset component detection model；The component
Detection model is based on being trained acquisition to initial part detection model using the first training sample data；In the first training sample
In data, sample characteristics include sample image, and sample label includes the component information of vehicle part in sample image.The portion of vehicle
Part size, component locations and component mark.
In model training stage, this specification embodiment is using sample image as sample characteristics, with vehicle in sample image
The component information of component constructs the first training sample data as sample label, and using training sample data to initial part
Detection model is trained, and obtains preset component detection model.It further, can also be using other component information as sample
Label, for example, it is also possible to component sizes, component compromise state, the component extent of damage, component are damaged size, component is damaged position
One of equal components relevant information or a variety of as sample label is set, so as to predict more using component detection model
Component information, with reach preferably guidance promoted effect.
The present embodiment training by the way of supervised learning obtains component detection model can adopt in one example
It uses deep learning model as initial part detection model, detects mould especially with MobileNets model construction initial part
Type ensure that model computational efficiency in the case where ensuring accuracy rate, can also accomplish to count in real time in the low sides type such as mobile terminal
Calculate feedback, it is possible to provide preferable user experience.
MobileNets is that the depth of lightweight is constructed using the separable convolution of depth based on a fairshaped framework
Layer neural network.By the global hyper parameter of introducing, balance is effectively performed between degree of delay and accuracy.Hyper parameter allows
Model construction person selects the model of suitable size for its application according to the constraint condition of problem.
Therefore, after obtaining component detection model, it can use the portion of target vehicle in component detection model forecast image
Part information improves the efficiency for obtaining component information.
Since the position of the component of target vehicle in image, size and title determine, then can predict filming apparatus with
The relative pose of target vehicle, for example, shooting distance and shooting angle.
About shooting distance information, the information for describing distance between photographing module and target vehicle can be.At one
In embodiment, shooting distance information can be distance value.In another embodiment, shooting distance information can be photographing module
With distance range affiliated at a distance from target vehicle.The distance relation of photographing module and target vehicle is described by distance range, with
It realizes and carries out relative efficiency can be improved, while reducing apart from detection difficulty apart from comparison using distance range.In an example
In, distance range can be embodied with digital scope.In another example, distance range can be embodied with distance level scale, for example,
By different distance range with it is remote, in, nearly three grades indicate, panorama, middle scape and close shot can be corresponded to.For example, when identified
When distance range and desired distance range difference, distance adjustment can be carried out and reminded.
In one embodiment, the shooting distance information is based on: with the output data of the component detection model and
Described image carries out prediction acquisition apart from detection model as the preset input data apart from detection model, and using described；
It is described to be based on being trained acquisition to initial distance detection model using the second training sample data apart from detection model；Second
In training sample data, sample characteristics include sample image, in sample image vehicle part component information, sample label includes
Shooting distance information.
In model training stage, this specification embodiment is with the component information of vehicle part in sample image, sample image
As sample characteristics, using shooting distance information as sample label, the second training sample data are constructed.And utilize the second training sample
Notebook data is trained initial distance detection model, obtains preset apart from detection model.The embodiment is not only by sample graph
As the prediction result apart from detection model can be improved also using component information as sample characteristics as sample characteristics.
The present embodiment training by the way of supervised learning is obtained apart from detection model, in one example, can be adopted
It uses deep learning model as initial distance detection model, detects mould especially with MobileNets model construction initial distance
Type.
In the application stage, the output data of component detection model and the preset distance of described image input can be detected
Model, using the data result apart from detection model as shooting distance information, realize using apart from detection model prediction shooting away from
From information, the accuracy of shooting distance information is improved.
About information of shooting angles, the mirror surface of photographing module and the relative angle information of target vehicle can be.At one
In embodiment, information of shooting angles can be specific shooting angle angle value.In another embodiment, information of shooting angles can be with
It is the affiliated angular range of shooting angle.For example, can with high angle shot, tiltedly clap, face upward bat, forehand etc. and indicate to take the photograph angle information.For another example, may be used
To indicate angular range with specific value range.
In one embodiment, the information of shooting angles is based on: with output data, the distance of the component detection model
Input data of the output data and described image of detection model as preset angle detection model, and utilize the angle
Detection model carries out prediction acquisition；The angle detection model is based on detecting mould to initial angle using third training sample data
Type is trained acquisition；In third training sample data, sample characteristics include sample image, vehicle part in sample image
Component information and shooting distance information, sample label include information of shooting angles.
In model training stage, this specification embodiment is with the component information of vehicle part in sample image, sample image
And shooting distance information, using information of shooting angles as sample label, constructs third training sample data as sample characteristics.
And initial distance detection model is trained using third training sample data, obtain preset angle detection model.The reality
Apply example not only using sample image as sample characteristics, it, can be with also using component information and shooting distance information as sample characteristics
Improve the prediction result of angle detection model.
The present embodiment training by the way of supervised learning obtains angle detection model can adopt in one example
It uses deep learning model as initial angle detection model, detects mould especially with MobileNets model construction initial angle
Type.
It, can be by the output data of component detection model, the output data apart from detection model and institute in the application stage
The angle detection model that image input training obtains is stated, it is real using the prediction result of angle detection model as information of shooting angles
Information of shooting angles now is predicted using angle detection model, improves the accuracy of information of shooting angles.
In one embodiment, the full link model of core all uses the deep learning model of mobile terminal, is ensuring accuracy rate
In the case where ensure that model computational efficiency, can also accomplish to calculate feedback in real time in low and middle-end type, it is possible to provide preferable to use
Family experience.
It is understood that in other embodiments, it can be by the bat of model estimation a photographing module and target vehicle
Photographic range information and information of shooting angles, will not repeat them here.
After determining current posture information based on acquired image, posture information and preset expected pose can be believed
Breath is compared, and is exported according to comparison result for guiding user's control photographing module with expected pose photographic subjects vehicle
Prompting message.Expected pose can be meet expected pose information in the case where opposite pass between photographing module and target vehicle
System.If shooting distance is not belonging to expectation shooting distance, control photographing module may remind the user that it is expected that shooting distance is shot
Target vehicle；If shooting angle is not belonging to expectation shooting angle, it may remind the user that control photographing module is clapped with expected angle
Take the photograph target vehicle etc..For example, may remind the user that control camera shooting model close to mesh if shooting distance is greater than desired shooting distance
Mark vehicle.For another example, if shooting distance information be " close ", and it is expected shooting distance information for " in ", then may remind the user that control
Camera shooting model departs slightly from target vehicle etc..
In one example, described shot with expected pose includes: close shot shooting, the shooting of middle scape, to specify shooting angle
Carry out one of pan-shot etc. or a variety of.Shooting distance corresponding to pan-shot is greater than the corresponding shooting of middle scape shooting
Distance, the corresponding shooting distance of middle scape shooting are greater than the corresponding shooting distance of close shot shooting.Specified shooting angle can be
45 degree of angles etc..It is defeated when determining and desired photographing information is not inconsistent in the shooting distance based on determined by current image and shooting angle
Corresponding prompting message out.For example, " 45 degree of vehicle panoramics please be shoot to shine according to comparison result output when executing vista shot
The prompting message of piece, and guarantee to see license plate "；It for another example, " please be close according to comparison result output when executing close shot shooting
It is some, shooting vehicle lose details, allow me to see the extent of damage clearly " prompting message；For another example, when scape is shot in commission, according to
Comparison result output " please retreat two steps, shoot vehicle damage position, me is allowed to see loss overview clearly ".
Wherein, the way of output of prompting message, can be voice broadcast, is also possible to text prompting, further, may be used also
To be that picture is reminded.For example, providing with the image of expected pose shooting sample vehicle, intuitively to be carried out to user using image
It reminds.
In order to guide user to shoot satisfactory car damage identification image, in one embodiment, can also scheme obtaining
Before picture, export with the indicator of expected pose photographic subjects vehicle.The purpose for exporting indicator is to remind user
The type of the image of required shooting, to obtain the car damage identification image of polymorphic type.For example, indicator may include that panorama is clapped
Take the photograph one of instruction, middle scape shooting instruction, close shot shooting instruction, specified angle shooting instruction etc. or a variety of.
As it can be seen that user may be implemented and shoot satisfactory multiclass under indicator prompting by output indicator
Type image.
Various technical characteristics in embodiment of above can be arbitrarily combined, as long as the combination between feature is not present
Conflict or contradiction, but as space is limited, it is not described one by one, therefore the various technical characteristics in above embodiment is any
It is combined the range for also belonging to this disclosure.
It is illustrated below with one of which combination.
It as shown in Figure 3A, is that this specification is shown according to an exemplary embodiment another for assisting car damage identification figure
As the process of the method for shooting.After picture shooting assembly is opened, photographing mode can be entered.In the mobile capture reality scene of camera lens
During, it can export with the indicator of expected pose photographic subjects vehicle, for example, indicator includes that pan-shot refers to
It enables, one of middle scape shooting instruction, close shot shooting instruction, specified angle shooting instruction or a variety of.Picture shooting assembly is by default frequency
Rate carries out cutting frame processing, and for every frame truncated picture, model built in invocation component is handled.Image input unit part is detected
Model obtains component information to carry out component identification to image.The output data of component detection model and described image is defeated
Enter it is preset apart from detection model, to obtain shooting distance information.By the output data of component detection model, apart from detection model
Output data and described image input training obtain angle detection model, to obtain information of shooting angles.According to component
Detection model, the output apart from detection module and angle detection model are as a result, export conclusion, picture shooting assembly root by Fusion Model
According to rule output user's prompt, user is prompted to need to adjust distance (adjust close or adjust remote) and shooting angle adjustment.In other implementations
In example, the image of acquisition can also be calculated in real time and generate prompt feedback.
It as shown in Figure 3B, is that this specification is shown according to an exemplary embodiment a kind of for assisting car damage identification image
The application example of shooting.Under every kind of shot type, exemplary diagram can be provided, for prompting user when not knowing how to shoot
It can be shot according to exemplary diagram.Fig. 3 B is illustrated with the instance graph in panorama.In this example, it is desirable that user claps
The image for taking the photograph three types, one is panoramic pictures, and one is close shot images, and one is middle scape images.For different images
Type configuration has corresponding expected pose information.User opens setting loss Baoying County and uses, into vehicle panoramic photographed scene.It is clapped in panorama
It takes the photograph in scene, if not meeting the expectation position of pan-shot based on the detected shooting distance of current interception image and shooting angle
When appearance information, exports prompting message: 45 degree of vehicle panoramic photos please be shoot, and guarantee to see license plate.Detect that user is mentioning
After showing the lower completion shooting action of guide, into scape photographed scene in vehicle.In middle scape photographed scene, if based on current interception figure
It when being less than the expectation shooting distance that middle scape is shot as detected shooting distance, exports prompting message: two steps please be retreat, shoot
Vehicle damage position allows me to see loss overview clearly.It is close into vehicle after detecting that user completes shooting action under prompt guide
Scape photographed scene.In entering close shot photographed scene, if being greater than close shot based on the detected shooting distance of current interception image
When the expectation shooting distance of shooting, export prompting message: please lean on it is closer, shooting vehicle lose details, allow me to see loss journey clearly
Degree.Detect that user after completing shooting action under prompt is guided, captured image can be shown, so that user is pre-
It lookes at, and executes setting loss image when submitting control by touch-control and submit operation.It is understood that knot can also be compared according to other
Fruit exports corresponding prompting message, will not repeat them here.As seen from the above-described embodiment, it according to the content of current shooting, utilizes
The AI model computing capability of mobile terminal, fusion component, distance, the calculated result of shooting angle export accurate shooting distance, bat
The position feedback informations such as angle are taken the photograph, so as to instruct user to adjust style of shooting, the more good car damage identification image of output.
Corresponding with the auxiliary embodiment of method of car damage identification image taking is previously used for, this specification additionally provides use
In the device of auxiliary car damage identification image taking and its embodiment of applied electronic equipment.
This specification is for assisting the embodiment of the device of car damage identification image taking that can apply in computer equipment.Dress
Setting embodiment can also be realized by software realization by way of hardware or software and hardware combining.It is implemented in software to be
Example, as the device on a logical meaning, being will be in nonvolatile memory by the processor of computer equipment where it
Corresponding computer program instructions are read into memory what operation was formed.For hardware view, as shown in figure 4, being this explanation
A kind of hardware structure diagram of computer equipment where book is used to assist the device of car damage identification image taking, in addition to shown in Fig. 4
Except processor 410, network interface 420, memory 430 and nonvolatile memory 440, for assisting vehicle in embodiment
Computer equipment where the device 431 of setting loss image taking can also include other generally according to the actual functional capability of the equipment
Hardware repeats no more this.
As shown in figure 5, being that this specification is shown according to an exemplary embodiment a kind of for assisting car damage identification image
The block diagram of the device of shooting, described device include:
Image collection module 52, is used for: obtaining the image of photographing module acquisition；
Information detecting module 54, is used for: the component of target vehicle in identification described image, and obtained according at least to identification
Component information detect photographing module and target vehicle relative pose, obtain posture information, the posture information include shooting away from
From one of information and information of shooting angles or a variety of；
Information reminding module 56, is used for: being obtained based on the posture information to be compared with preset expected pose information
The comparison result obtained is exported for guiding user's control photographing module with the prompting message of expected pose photographic subjects vehicle.
In one embodiment, the component information include: component locations, component sizes and component mark, and/or, institute
Stating shooting distance information is photographing module affiliated distance range at a distance from target vehicle, and/or, the information of shooting angles is
The affiliated angular range of shooting angle.
In one embodiment, the component information is based on knowing described image using preset component detection model
It does not obtain；The component detection model is based on being trained initial part detection model using the first training sample data obtaining
?；In the first training sample data, sample characteristics include sample image, and sample label includes vehicle part in sample image
Component information.
In one embodiment, the shooting distance information is based on: with the output data of the component detection model and
Described image carries out prediction acquisition apart from detection model as the preset input data apart from detection model, and using described；
It is described to be based on being trained acquisition to initial distance detection model using the second training sample data apart from detection model；Second
In training sample data, sample characteristics include sample image, in sample image vehicle part component information, sample label includes
Shooting distance information.
In one embodiment, the information of shooting angles is based on: with output data, the distance of the component detection model
Input data of the output data and described image of detection model as preset angle detection model, and utilize the angle
Detection model carries out prediction acquisition；The angle detection model is based on detecting mould to initial angle using third training sample data
Type is trained acquisition；In third training sample data, sample characteristics include sample image, vehicle part in sample image
Component information and shooting distance information, sample label include information of shooting angles.
In one embodiment, the initial part detection model, the initial distance detection model and initial angle
Detection model respectively includes MobileNets model.
In one embodiment, described shot with expected pose includes: to carry out pan-shot, close shot with specified shooting angle
Shooting, middle scape shooting etc. one of or it is a variety of, the pan-shot, middle scape shooting and close shot shooting by shooting distance from greatly to
Small sequence is divided.
For device embodiment, since it corresponds essentially to embodiment of the method, so related place is referring to method reality
Apply the part explanation of example.The apparatus embodiments described above are merely exemplary, wherein described be used as separation unit
The module of explanation may or may not be physically separated, and the component shown as module can be or can also be with
It is not physical module, it can it is in one place, or may be distributed on multiple network modules.It can be according to actual
The purpose for needing to select some or all of the modules therein to realize this specification scheme.Those of ordinary skill in the art are not
In the case where making the creative labor, it can understand and implement.
Correspondingly, this specification embodiment also provides a kind of computer equipment, including memory, processor and it is stored in
On reservoir and the computer program that can run on a processor, wherein the processor realizes such as lower section when executing described program
Method:
Obtain the image of photographing module acquisition；
Identify the component of target vehicle in described image, and the component information obtained according at least to identification detects photographing module
With the relative pose of target vehicle, posture information is obtained, the posture information includes shooting distance information and information of shooting angles
One of or it is a variety of；
Based on the comparison result that the posture information is compared to acquisition with preset expected pose information, output is used for
Guide user's control photographing module with the prompting message of expected pose photographic subjects vehicle.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for equipment reality
For applying example, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring to embodiment of the method
Part explanation.
A kind of computer storage medium, program instruction is stored in the storage medium, and described program instruction includes:
Obtain the image of photographing module acquisition；
Identify the component of target vehicle in described image, and the component information obtained according at least to identification detects photographing module
With the relative pose of target vehicle, posture information is obtained, the posture information includes shooting distance information and information of shooting angles
One of or it is a variety of；
Based on the comparison result that the posture information is compared to acquisition with preset expected pose information, output is used for
Guide user's control photographing module with the prompting message of expected pose photographic subjects vehicle.
This specification embodiment can be used one or more wherein include the storage medium of program code (including but not
Be limited to magnetic disk storage, CD-ROM, optical memory etc.) on the form of computer program product implemented.Computer is available to be deposited
Storage media includes permanent and non-permanent, removable and non-removable media, can be accomplished by any method or technique letter
Breath storage.Information can be computer readable instructions, data structure, the module of program or other data.The storage of computer is situated between
The example of matter includes but is not limited to: phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory
Device (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), the read-only storage of electrically erasable
Device (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM), digital versatile disc
(DVD) or other optical storage, magnetic cassettes, tape magnetic disk storage or other magnetic storage devices or any other non-biography
Defeated medium, can be used for storage can be accessed by a computing device information.
Those skilled in the art will readily occur to this specification after considering specification and practicing the invention applied here
Other embodiments.This specification is intended to cover any variations, uses, or adaptations of this specification, these modifications,
Purposes or adaptive change follow the general principle of this specification and do not apply in the art including this specification
Common knowledge or conventional techniques.The description and examples are only to be considered as illustrative, the true scope of this specification and
Spirit is indicated by the following claims.
It should be understood that this specification is not limited to the precise structure that has been described above and shown in the drawings,
And various modifications and changes may be made without departing from the scope thereof.The range of this specification is only limited by the attached claims
System.
The foregoing is merely the preferred embodiments of this specification, all in this explanation not to limit this specification
Within the spirit and principle of book, any modification, equivalent substitution, improvement and etc. done should be included in the model of this specification protection
Within enclosing.
Claims (12)
1. a kind of method for assisting car damage identification image taking, which comprises
Obtain the image of photographing module acquisition；
Identify the component of target vehicle in described image, and according at least to the component information detection photographing module and mesh that identification obtains
The relative pose of vehicle is marked, obtains posture information, the posture information includes in shooting distance information and information of shooting angles
It is one or more；
Based on the comparison result that the posture information is compared to acquisition with preset expected pose information, export for guiding
User's control photographing module is with the prompting message of expected pose photographic subjects vehicle.
2. according to the method described in claim 1, the component information include: component locations, component sizes and component mark,
And/or the shooting distance information is photographing module affiliated distance range at a distance from target vehicle, and/or, the shooting angle
Degree information is the affiliated angular range of shooting angle.
3. method according to claim 1 or 2, the component information is based on using preset component detection model to described
Image carries out identification acquisition；The component detection model be based on using the first training sample data to initial part detection model into
Row training obtains；In the first training sample data, sample characteristics include sample image, and sample label includes vehicle in sample image
The component information of component.
4. according to the method described in claim 3, the shooting distance information is based on: with the output number of the component detection model
Accordingly and described image is predicted as the preset input data apart from detection model, and using described apart from detection model
It obtains；It is described to be based on being trained acquisition to initial distance detection model using the second training sample data apart from detection model；
In the second training sample data, sample characteristics include sample image, in sample image vehicle part component information, sample mark
Label include shooting distance information.
5. according to the method described in claim 4, the information of shooting angles is based on: with the output number of the component detection model
According to, the input data of the output data apart from detection model and described image as preset angle detection model, and utilize
The angle detection model carries out prediction acquisition；The angle detection model is based on using third training sample data to initial angle
Degree detection model is trained acquisition；In third training sample data, sample characteristics include sample image, vehicle in sample image
The component information and shooting distance information of component, sample label includes information of shooting angles.
6. according to the method described in claim 5, the initial part detection model, the initial distance detection model and just
Beginning angle detection model respectively includes MobileNets model.
7. method according to any one of claims 1 to 6, described shot with expected pose includes: to specify shooting angle
One of pan-shot, close shot shooting, the shooting of middle scape etc. or a variety of are carried out, the pan-shot, the shooting of middle scape and close shot are clapped
It takes the photograph and is divided by the sequence of shooting distance from big to small.
8. a kind of for assisting the device of car damage identification image taking, described device includes:
Image collection module is used for: obtaining the image of photographing module acquisition；
Information detecting module is used for: the component of target vehicle in identification described image, and the component obtained according at least to identification is believed
The relative pose of breath detection photographing module and target vehicle, obtains posture information, the posture information includes shooting distance information
With one of information of shooting angles or a variety of；
Information reminding module, is used for: based on the ratio that the posture information and preset expected pose information are compared to acquisition
Compared with as a result, exporting for guiding user's control photographing module with the prompting message of expected pose photographic subjects vehicle.
9. device according to claim 8, the component information is based on using preset component detection model to the figure
As carrying out identification acquisition；The component detection model is based on carrying out initial part detection model using the first training sample data
Training obtains；In the first training sample data, sample characteristics include sample image, and sample label includes vehicle in sample image
The component information of component.
10. device according to claim 9, the shooting distance information is based on: with the output of the component detection model
Data and described image are carried out in advance as the preset input data apart from detection model, and using described apart from detection model
It surveys and obtains；It is described to be based on being trained initial distance detection model using the second training sample data obtaining apart from detection model
?；In the second training sample data, sample characteristics include sample image, in sample image vehicle part component information, sample
This label includes shooting distance information.
11. device according to claim 10, the information of shooting angles is based on: with the output of the component detection model
The input data of data, the output data apart from detection model and described image as preset angle detection model, and benefit
Prediction acquisition is carried out with the angle detection model；The angle detection model is based on using third training sample data to initial
Angle detection model is trained acquisition；In third training sample data, sample characteristics include sample image, in sample image
The component information and shooting distance information of vehicle part, sample label includes information of shooting angles.
12. a kind of computer equipment including memory, processor and stores the meter that can be run on a memory and on a processor
Calculation machine program, wherein the processor realizes following method when executing described program:
Obtain the image of photographing module acquisition；
Identify the component of target vehicle in described image, and according at least to the component information detection photographing module and mesh that identification obtains
The relative pose of vehicle is marked, obtains posture information, the posture information includes in shooting distance information and information of shooting angles
It is one or more；
Based on the comparison result that the posture information is compared to acquisition with preset expected pose information, export for guiding
User's control photographing module is with the prompting message of expected pose photographic subjects vehicle.
Priority Applications (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811013914.6A CN109325488A (en) | 2018-08-31 | 2018-08-31 | For assisting the method, device and equipment of car damage identification image taking |
| TW108122283A TWI710967B (en) | 2018-08-31 | 2019-06-26 | Method, device and equipment for assisting vehicle damage fixing image shooting |
| PCT/CN2019/096321 WO2020042800A1 (en) | 2018-08-31 | 2019-07-17 | Auxiliary method for capturing damage assessment image of vehicle, device, and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811013914.6A CN109325488A (en) | 2018-08-31 | 2018-08-31 | For assisting the method, device and equipment of car damage identification image taking |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109325488A true CN109325488A (en) | 2019-02-12 |
Family
ID=65264238
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811013914.6A Pending CN109325488A (en) | 2018-08-31 | 2018-08-31 | For assisting the method, device and equipment of car damage identification image taking |
Country Status (3)
| Country | Link |
|---|---|
| CN (1) | CN109325488A (en) |
| TW (1) | TWI710967B (en) |
| WO (1) | WO2020042800A1 (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110033386A (en) * | 2019-03-07 | 2019-07-19 | 阿里巴巴集团控股有限公司 | Identification method and device, the electronic equipment of car accident |
| CN110245552A (en) * | 2019-04-29 | 2019-09-17 | 阿里巴巴集团控股有限公司 | Interaction processing method, device, equipment and the client of vehicle damage image taking |
| CN110264444A (en) * | 2019-05-27 | 2019-09-20 | 阿里巴巴集团控股有限公司 | Damage detecting method and device based on weak segmentation |
| CN110598562A (en) * | 2019-08-15 | 2019-12-20 | 阿里巴巴集团控股有限公司 | Vehicle image acquisition guiding method and device |
| CN110659568A (en) * | 2019-08-15 | 2020-01-07 | 阿里巴巴集团控股有限公司 | Vehicle inspection method and device |
| CN110659567A (en) * | 2019-08-15 | 2020-01-07 | 阿里巴巴集团控股有限公司 | Method and device for identifying damaged part of vehicle |
| CN110658731A (en) * | 2019-10-21 | 2020-01-07 | 珠海格力电器股份有限公司 | Intelligent household appliance network distribution method, storage medium and intelligent terminal |
| WO2020042800A1 (en) * | 2018-08-31 | 2020-03-05 | 阿里巴巴集团控股有限公司 | Auxiliary method for capturing damage assessment image of vehicle, device, and apparatus |
| CN110910628A (en) * | 2019-12-02 | 2020-03-24 | 支付宝(杭州)信息技术有限公司 | Interactive processing method and device for vehicle damage image shooting and electronic equipment |
| US10783643B1 (en) | 2019-05-27 | 2020-09-22 | Alibaba Group Holding Limited | Segmentation-based damage detection |
| CN111985448A (en) * | 2020-09-02 | 2020-11-24 | 深圳壹账通智能科技有限公司 | Vehicle image recognition method and device, computer equipment and readable storage medium |
| CN112288800A (en) * | 2020-09-27 | 2021-01-29 | 济南浪潮高新科技投资发展有限公司 | Server cabinet door eye identification method, equipment and device |
| CN112348686A (en) * | 2020-11-24 | 2021-02-09 | 德联易控科技(北京)有限公司 | Claim settlement picture acquisition method and device and communication equipment |
| CN112364820A (en) * | 2020-11-27 | 2021-02-12 | 深源恒际科技有限公司 | Deep learning-based vehicle insurance underwriting and vehicle checking picture acquisition method and system |
| WO2021047249A1 (en) * | 2019-09-09 | 2021-03-18 | 平安科技（深圳）有限公司 | Data prediction method, apparatus and device, and computer-readable storage medium |
| CN112585655A (en) * | 2020-11-23 | 2021-03-30 | 北京零能动力汽车科技有限公司 | Unmanned electronic traffic police duty system based on 5G |
| CN113038018A (en) * | 2019-10-30 | 2021-06-25 | 支付宝(杭州)信息技术有限公司 | Method and device for assisting user in shooting vehicle video |
| CN113132632A (en) * | 2021-04-06 | 2021-07-16 | 支付宝(杭州)信息技术有限公司 | Auxiliary shooting method and device for pets |
| CN113627252A (en) * | 2021-07-07 | 2021-11-09 | 浙江吉利控股集团有限公司 | Vehicle damage assessment method and device, storage medium and electronic equipment |
| CN113840085A (en) * | 2021-09-02 | 2021-12-24 | 北京城市网邻信息技术有限公司 | Vehicle source information acquisition method and device, electronic equipment and readable medium |
| CN117455466A (en) * | 2023-12-22 | 2024-01-26 | 南京三百云信息科技有限公司 | Method and system for remote evaluation of automobile |
| CN118430818A (en) * | 2024-07-04 | 2024-08-02 | 山东大学齐鲁医院 | Endoscopic gastric cancer risk classification system, medium and equipment based on artificial intelligence |
| WO2024174718A1 (en) * | 2023-02-24 | 2024-08-29 | 华为技术有限公司 | Image acquisition guiding method and apparatus, and vehicle |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111553268A (en) * | 2020-04-27 | 2020-08-18 | 深圳壹账通智能科技有限公司 | Vehicle part identification method and device, computer equipment and storage medium |
| CN111918049B (en) * | 2020-08-14 | 2022-09-06 | 广东申义实业投资有限公司 | Three-dimensional imaging method and device, electronic equipment and storage medium |
| CN112492105B (en) * | 2020-11-26 | 2022-04-15 | 深源恒际科技有限公司 | Video-based vehicle appearance part self-service damage assessment acquisition method and system |
| CN112633295B (en) * | 2020-12-22 | 2024-06-18 | 深圳集智数字科技有限公司 | Prediction method and device for cyclic task, electronic equipment and storage medium |
| CN113516013B (en) * | 2021-04-09 | 2024-05-14 | 阿波罗智联(北京)科技有限公司 | Target detection method, target detection device, electronic equipment, road side equipment and cloud control platform |
| CN114219806B (en) * | 2022-02-22 | 2022-04-22 | 成都数联云算科技有限公司 | Automobile radar detection method, device, equipment, medium and product |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016132251A1 (en) * | 2015-02-16 | 2016-08-25 | Coolbox S.R.L. | Device and method for monitoring a vehicle, particularly for the management of loss events |
| CN106139564A (en) * | 2016-08-01 | 2016-11-23 | 纳恩博（北京）科技有限公司 | Image processing method and device |
| CN106600421A (en) * | 2016-11-21 | 2017-04-26 | 中国平安财产保险股份有限公司 | Intelligent car insurance loss assessment method and system based on image recognition |
| CN106657755A (en) * | 2015-07-30 | 2017-05-10 | 中兴通讯股份有限公司 | Photographing method and photographing device |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9232143B2 (en) * | 2013-09-18 | 2016-01-05 | Wipro Limited | Method and system for assistive photography |
| CN104407351B (en) * | 2014-12-05 | 2017-05-17 | 北京公科飞达交通工程发展有限公司 | Vehicle window position identification method |
| CN104637342B (en) * | 2015-01-22 | 2017-01-04 | 江苏大学 | A kind of narrow and small vertical parking stall scene intelligent identification and parking path planning system and method |
| CN105069451B (en) * | 2015-07-08 | 2018-05-25 | 北京智能综电信息技术有限责任公司 | A kind of Car license recognition and localization method based on binocular camera |
| CN107229625A (en) * | 2016-03-23 | 2017-10-03 | 北京搜狗科技发展有限公司 | It is a kind of to shoot treating method and apparatus, a kind of device for being used to shoot processing |
| TWI603271B (en) * | 2016-10-20 | 2017-10-21 | 元智大學 | Automatic electric scooter identification an d part-based outer defect detection method and system thereof |
| CN108240994A (en) * | 2016-12-24 | 2018-07-03 | 天津科寻科技有限公司 | The device whether scratched using internet of things sensors monitoring automobile chassis |
| CN109325488A (en) * | 2018-08-31 | 2019-02-12 | 阿里巴巴集团控股有限公司 | For assisting the method, device and equipment of car damage identification image taking |
-
2018
- 2018-08-31 CN CN201811013914.6A patent/CN109325488A/en active Pending
-
2019
- 2019-06-26 TW TW108122283A patent/TWI710967B/en active
- 2019-07-17 WO PCT/CN2019/096321 patent/WO2020042800A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016132251A1 (en) * | 2015-02-16 | 2016-08-25 | Coolbox S.R.L. | Device and method for monitoring a vehicle, particularly for the management of loss events |
| CN106657755A (en) * | 2015-07-30 | 2017-05-10 | 中兴通讯股份有限公司 | Photographing method and photographing device |
| CN106139564A (en) * | 2016-08-01 | 2016-11-23 | 纳恩博（北京）科技有限公司 | Image processing method and device |
| CN106600421A (en) * | 2016-11-21 | 2017-04-26 | 中国平安财产保险股份有限公司 | Intelligent car insurance loss assessment method and system based on image recognition |
Cited By (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020042800A1 (en) * | 2018-08-31 | 2020-03-05 | 阿里巴巴集团控股有限公司 | Auxiliary method for capturing damage assessment image of vehicle, device, and apparatus |
| CN110033386A (en) * | 2019-03-07 | 2019-07-19 | 阿里巴巴集团控股有限公司 | Identification method and device, the electronic equipment of car accident |
| CN110033386B (en) * | 2019-03-07 | 2020-10-02 | 阿里巴巴集团控股有限公司 | Vehicle accident identification method and device and electronic equipment |
| CN110245552A (en) * | 2019-04-29 | 2019-09-17 | 阿里巴巴集团控股有限公司 | Interaction processing method, device, equipment and the client of vehicle damage image taking |
| CN110245552B (en) * | 2019-04-29 | 2023-07-18 | 创新先进技术有限公司 | Interactive processing method, device, equipment and client for vehicle damage image shooting |
| US11004204B2 (en) | 2019-05-27 | 2021-05-11 | Advanced New Technologies Co., Ltd. | Segmentation-based damage detection |
| CN110264444A (en) * | 2019-05-27 | 2019-09-20 | 阿里巴巴集团控股有限公司 | Damage detecting method and device based on weak segmentation |
| US10783643B1 (en) | 2019-05-27 | 2020-09-22 | Alibaba Group Holding Limited | Segmentation-based damage detection |
| CN110264444B (en) * | 2019-05-27 | 2020-07-17 | 阿里巴巴集团控股有限公司 | Damage detection method and device based on weak segmentation |
| CN110598562A (en) * | 2019-08-15 | 2019-12-20 | 阿里巴巴集团控股有限公司 | Vehicle image acquisition guiding method and device |
| CN110598562B (en) * | 2019-08-15 | 2023-03-07 | 创新先进技术有限公司 | Vehicle image acquisition guiding method and device |
| CN110659567A (en) * | 2019-08-15 | 2020-01-07 | 阿里巴巴集团控股有限公司 | Method and device for identifying damaged part of vehicle |
| CN110659568A (en) * | 2019-08-15 | 2020-01-07 | 阿里巴巴集团控股有限公司 | Vehicle inspection method and device |
| WO2021047249A1 (en) * | 2019-09-09 | 2021-03-18 | 平安科技（深圳）有限公司 | Data prediction method, apparatus and device, and computer-readable storage medium |
| CN110658731A (en) * | 2019-10-21 | 2020-01-07 | 珠海格力电器股份有限公司 | Intelligent household appliance network distribution method, storage medium and intelligent terminal |
| CN110658731B (en) * | 2019-10-21 | 2020-07-03 | 珠海格力电器股份有限公司 | Intelligent household appliance network distribution method, storage medium and intelligent terminal |
| CN113038018A (en) * | 2019-10-30 | 2021-06-25 | 支付宝(杭州)信息技术有限公司 | Method and device for assisting user in shooting vehicle video |
| CN110910628A (en) * | 2019-12-02 | 2020-03-24 | 支付宝(杭州)信息技术有限公司 | Interactive processing method and device for vehicle damage image shooting and electronic equipment |
| CN111985448A (en) * | 2020-09-02 | 2020-11-24 | 深圳壹账通智能科技有限公司 | Vehicle image recognition method and device, computer equipment and readable storage medium |
| CN112288800A (en) * | 2020-09-27 | 2021-01-29 | 济南浪潮高新科技投资发展有限公司 | Server cabinet door eye identification method, equipment and device |
| CN112288800B (en) * | 2020-09-27 | 2023-05-12 | 山东浪潮科学研究院有限公司 | Method, equipment and device for identifying door lock eyes of server cabinet |
| CN112585655A (en) * | 2020-11-23 | 2021-03-30 | 北京零能动力汽车科技有限公司 | Unmanned electronic traffic police duty system based on 5G |
| CN112348686A (en) * | 2020-11-24 | 2021-02-09 | 德联易控科技(北京)有限公司 | Claim settlement picture acquisition method and device and communication equipment |
| CN112348686B (en) * | 2020-11-24 | 2021-07-13 | 德联易控科技(北京)有限公司 | Claim settlement picture acquisition method and device and communication equipment |
| CN112364820A (en) * | 2020-11-27 | 2021-02-12 | 深源恒际科技有限公司 | Deep learning-based vehicle insurance underwriting and vehicle checking picture acquisition method and system |
| WO2022213974A1 (en) * | 2021-04-06 | 2022-10-13 | 蚂蚁胜信(上海)信息技术有限公司 | Auxiliary image capture methods and apparatuses for pets |
| CN113132632A (en) * | 2021-04-06 | 2021-07-16 | 支付宝(杭州)信息技术有限公司 | Auxiliary shooting method and device for pets |
| EP4322515A4 (en) * | 2021-04-06 | 2024-09-18 | Ant Shengxin Shanghai Information Tech Co Ltd | Auxiliary image capture methods and apparatuses for pets |
| CN113627252A (en) * | 2021-07-07 | 2021-11-09 | 浙江吉利控股集团有限公司 | Vehicle damage assessment method and device, storage medium and electronic equipment |
| CN113840085A (en) * | 2021-09-02 | 2021-12-24 | 北京城市网邻信息技术有限公司 | Vehicle source information acquisition method and device, electronic equipment and readable medium |
| WO2024174718A1 (en) * | 2023-02-24 | 2024-08-29 | 华为技术有限公司 | Image acquisition guiding method and apparatus, and vehicle |
| CN117455466A (en) * | 2023-12-22 | 2024-01-26 | 南京三百云信息科技有限公司 | Method and system for remote evaluation of automobile |
| CN117455466B (en) * | 2023-12-22 | 2024-03-08 | 南京三百云信息科技有限公司 | Method and system for remote evaluation of automobile |
| CN118430818A (en) * | 2024-07-04 | 2024-08-02 | 山东大学齐鲁医院 | Endoscopic gastric cancer risk classification system, medium and equipment based on artificial intelligence |
| CN118430818B (en) * | 2024-07-04 | 2024-10-22 | 山东大学齐鲁医院 | Endoscopic gastric cancer risk classification system, medium and equipment based on artificial intelligence |
Also Published As
| Publication number | Publication date |
|---|---|
| TW202011254A (en) | 2020-03-16 |
| WO2020042800A1 (en) | 2020-03-05 |
| TWI710967B (en) | 2020-11-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109325488A (en) | For assisting the method, device and equipment of car damage identification image taking | |
| TWI709091B (en) | Image processing method and device | |
| EP3605386B1 (en) | Method and apparatus for obtaining vehicle loss assessment image, server and terminal device | |
| CN108710847B (en) | Scene recognition method and device and electronic equipment | |
| CN107481327B (en) | About the processing method of augmented reality scene, device, terminal device and system | |
| WO2018036276A1 (en) | Image quality detection method, device, server and storage medium | |
| CN109815843B (en) | Image processing method and related product | |
| KR20190069457A (en) | IMAGE BASED VEHICLES LOSS EVALUATION METHOD, DEVICE AND SYSTEM, | |
| WO2021082662A1 (en) | Method and apparatus for assisting user in shooting vehicle video | |
| CN110163053A (en) | Generate the method, apparatus and computer equipment of the negative sample of recognition of face | |
| CN110910628B (en) | Interactive processing method and device for vehicle damage image shooting and electronic equipment | |
| WO2018192245A1 (en) | Automatic scoring method for photo based on aesthetic assessment | |
| CN110263615A (en) | Interaction processing method, device, equipment and client in vehicle shooting | |
| CN108182746A (en) | Control system, method and apparatus | |
| CN111881764A (en) | Target detection method and device, electronic equipment and storage medium | |
| CN110766077A (en) | Method, device and equipment for screening sketch in evidence chain image | |
| CN110427998A (en) | Model training, object detection method and device, electronic equipment, storage medium | |
| CN111881740A (en) | Face recognition method, face recognition device, electronic equipment and medium | |
| CN115880695A (en) | Card identification method, card identification model training method and electronic equipment | |
| CN112418303B (en) | Training method and device for identifying state model and computer equipment | |
| WO2018192244A1 (en) | Shooting guidance method for intelligent device | |
| CN110047115B (en) | Star image shooting method and device, computer equipment and storage medium | |
| CN109298783A (en) | Mark monitoring method, device and electronic equipment based on Expression Recognition | |
| CN113807150A (en) | Data processing method, attitude prediction method, data processing device, attitude prediction device, and storage medium | |
| AU2023213179A1 (en) | Crowdsourcing techniques to deploy artificial intelligence systems |
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 | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 40004213 Country of ref document: HK |
|
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190212 |
|
| RJ01 | Rejection of invention patent application after publication |