CN111881856A - Vehicle damage assessment method and device based on image - Google Patents
Vehicle damage assessment method and device based on image Download PDFInfo
- Publication number
- CN111881856A CN111881856A CN202010756372.2A CN202010756372A CN111881856A CN 111881856 A CN111881856 A CN 111881856A CN 202010756372 A CN202010756372 A CN 202010756372A CN 111881856 A CN111881856 A CN 111881856A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- damage
- component
- images
- impairment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
- G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Abstract
The invention relates to a vehicle damage assessment method and device based on images. Concretely, a vehicle damage assessment method is provided, comprising: acquiring a vehicle damage image set of a vehicle; determining a component configuration type for the vehicle, the component configuration type for the vehicle corresponding to a component configuration for the vehicle, the component configuration comprising a mapping of each exterior component of the vehicle to one or more interior components; and inputting the component configuration type of the vehicle and the vehicle damage image set into a neural network model to determine the internal damage component of the vehicle and the damage probability thereof.
Description
Technical Field
The application relates to the technical field of computer image processing, in particular to a vehicle damage assessment method and device based on images.
Background
After a traffic accident occurs, a claimant of an insurance company is often required to arrive at a field for processing, and a claim settlement basis is obtained by taking a picture and the like. With the increase of the keeping quantity of motor vehicles in recent years, the processing of vehicle claim settlement and damage service needs the manpower field processing depending on professional insurance payers, and has high cost, long waiting period and low processing efficiency.
Disclosure of Invention
In order to solve the technical problem, the disclosure provides a full-link automatic claims settlement scheme for car insurance, so that a user can quickly and conveniently complete all processes of claims settlement.
The present disclosure provides a vehicle damage assessment method, comprising:
acquiring a vehicle damage image set of a vehicle;
determining a component configuration type for the vehicle, the component configuration type for the vehicle corresponding to a component configuration for the vehicle, the component configuration comprising a mapping of each exterior component of the vehicle to one or more interior components; and
inputting the component configuration type of the vehicle and the vehicle damage image set into a neural network model to determine internal damaged components of the vehicle and damage probabilities thereof.
Optionally, the method further comprises:
processing the set of impairment images to determine an external impairment component of the vehicle; and
determining an internal damaged component of the vehicle and a damage probability thereof according to the component configuration type of the vehicle, the external damaged component and the vehicle damage image set.
Optionally, the method further comprises:
processing the set of vehicle damage images to determine an external damaged component of the vehicle and its degree of damage; and
and determining a maintenance scheme of the vehicle according to the external damaged component and the damage degree thereof, and the internal damaged component and the damage probability thereof.
Optionally, the component configuration type of the vehicle is determined by a vehicle identification code.
Optionally, the method further comprises: determining that the internal damaged component is to be claim if the probability of damage to the internal damaged component is above a threshold.
Optionally, the method as recited in claim 1, further comprising:
training the neural network model by using a training sample, wherein the training sample comprises a vehicle damage image set of the determined damage vehicle insurance, a component configuration type of the vehicle and an identification number of an internal damage component.
Optionally, the method further comprises:
augmenting a set of images of the vehicle, the augmenting including rotating, translating, scaling, and color transforming images of the set of images.
Optionally, the method further comprises:
determining a quality of each image in the set of impairment images; and
and screening the images in the vehicle damage image set according to the quality of the images.
Another aspect of the present disclosure provides a vehicle damage assessment apparatus, including:
means for obtaining a set of impairment images for a vehicle;
means for determining a component configuration type of the vehicle, the component configuration type of the vehicle corresponding to a component configuration of the vehicle, the component configuration comprising a mapping of each exterior component of the vehicle to one or more interior components; and
means for inputting the component configuration type of the vehicle and the set of vehicle damage images into a neural network model to determine an interior damaged component of the vehicle and its damage probability.
Optionally, the apparatus further comprises:
means for processing the set of impairment images to determine an external impairment component of the vehicle; and
means for determining an interior damaged component of the vehicle and its damage probability from a component configuration type of the vehicle, the exterior damaged component, and the impairment image set.
Optionally, the apparatus further comprises:
means for processing the set of impairment images to determine an external impairment component of the vehicle and its extent of impairment; and
means for determining a repair scenario for the vehicle based on the externally damaged component and its degree of damage, and the internally damaged component and its probability of damage.
Optionally, the component configuration type of the vehicle is determined by a vehicle identification code.
Optionally, the apparatus further comprises:
means for determining that the internal damaged component is to be claimed if the probability of damage to the internal damaged component is above a threshold.
Optionally, the apparatus further comprises:
a module for training the neural network model using training samples, the training samples including a set of impairment images of an impairment car risk, a component configuration type of a vehicle, an identification number of an internal impairment component.
Optionally, the apparatus further comprises:
means for augmenting a set of images of the vehicle, the augmenting including rotating, translating, scaling, and color transforming images of the set of images.
Optionally, the apparatus further comprises:
means for determining a quality of each image in the set of impairment images; and
and the module is used for screening the images in the vehicle damage image set according to the quality of the images.
Yet another aspect of the present disclosure provides a vehicle damage assessment apparatus, comprising:
a processor; and
a memory arranged to store computer executable instructions that, when executed, cause the processor to:
acquiring a vehicle damage image set of a vehicle;
determining a component configuration type for the vehicle, the component configuration type for the vehicle corresponding to a component configuration for the vehicle, the component configuration comprising a mapping of each exterior component of the vehicle to one or more interior components; and
inputting the component configuration type of the vehicle and the vehicle damage image set into a neural network model to determine internal damaged components of the vehicle and damage probabilities thereof.
According to the technical scheme, the damage condition of the external part and the damage condition of the internal part of the vehicle can be determined based on the vehicle damage image set provided by the user and the component configuration type of the vehicle and the constructed damage identification model, the maintenance scheme and the estimated claim settlement price of the vehicle are further automatically determined, manual intervention is not needed in the processing process, and the full-link automatic claim settlement scheme of the vehicle insurance is realized.
Drawings
Fig. 1 is a diagram of a system for image-based vehicle damage assessment according to various aspects of the present disclosure.
Fig. 2 is a flow diagram of a method for image-based vehicle damage assessment according to aspects of the present disclosure.
Fig. 3 is a diagram of an apparatus for predicting vehicle damage according to an aspect of the present disclosure.
FIG. 4 is a flow diagram of a method of training a neural network model for vehicle impairment in accordance with aspects of the present disclosure.
Fig. 5 is a diagram of an apparatus for image-based vehicle damage assessment according to aspects of the present disclosure.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways than those specifically described herein, and thus the present invention is not limited to the specific embodiments disclosed below.
In the prior art, damage and damage degree of external components of a vehicle (i.e., components visible from the outside of the vehicle, such as a bumper skin, an engine hood, etc.) can be determined through images uploaded by a user, but professional maintenance personnel are still required to determine damage of internal components of the vehicle (i.e., components not visible from the outside of the vehicle, such as a bumper frame, a radiator, a fan, etc.), which wastes a lot of human resources, prolongs a claim period, and also reduces user experience.
The vehicle damage assessment method based on the images aims at the problems, the improved vehicle damage assessment scheme based on the images is provided, and the damage probability of relevant internal components is predicted according to the external vehicle damage images of the vehicle, so that the vehicle damage assessment process can be automated, human intervention is not needed, human resources are saved, the claim settlement period is shortened, and the user experience is improved.
Fig. 1 is a diagram of a system for image-based vehicle damage assessment according to various aspects of the present disclosure.
As shown in FIG. 1, the system 100 may include a plurality of terminals 1011-N, a server 102, and a database 103. The plurality of terminals 101-1-N, the server 102 and the database 103 may communicate via a wired or wireless connection.
The terminal 101 may be a device having a network connection function. The terminal 101 may be installed with a car insurance claim settlement application client. The terminal 101 may include a camera to capture images (e.g., a loss of car image). The car insurance claims settlement application client can guide the user to shoot car damage images. The vehicle damage image may include a full vehicle image including license plate information, a distant view image of a vehicle damage portion, a close view image of the vehicle damage portion, and a vehicle damage detail image. If the vehicle of the user has more than one damage, the user can be guided to respectively shoot a distant view image of the vehicle damage part, a close view image of the vehicle damage part and a vehicle damage detail image aiming at each damage so as to completely record all vehicle damage information. The terminal 101 can also take pictures of vehicle certificates and user certificates. In this context, a set of impairment images includes images related to vehicle impairment, including vehicle images and related credential images. The terminal 101 may also receive relevant information input by the user.
The terminal 101 may send the set of impairment images and other user input information to the server 102 for processing. The terminal 101 may then receive the damage assessment information from the server 102 for a claim validation operation.
The server 102 may be a single server or a server cluster including a plurality of servers. The server 102 may provide various business services to a plurality of terminals 1011-N.
The server 102 may receive the set of loss images from the terminal 101, process the set of loss images to extract relevant information, such as a vehicle identification code, user identity information, and the like. The server 102 may also filter the collection of impairment images, e.g., filter out unsatisfactory images based on whether the images are of a quality requirement (e.g., whether the images are sufficiently sharp, whether the impairment site is captured, etc.), and may notify the terminal 101 to re-capture the corresponding images.
The server 102 may determine the component configuration type of the vehicle. For example, the server 102 may determine the model, the factory time, and the configuration level (e.g., high, medium, low) of the vehicle based on the identification code of the vehicle. The identification code of the vehicle can be obtained from the corresponding image in the vehicle damage image set, and can also be input by a user.
Vehicles of different models, delivery times, and configuration levels may have different component configurations. Herein, component configuration refers to a map of damage to an exterior component to one or more interior components due to the placement of the components of the vehicle. For example, in one component configuration, a particular outer component a may be adjacent to inner component B and inner component C, whereby damage to the particular outer component a may cause damage to inner component B and inner component C; in yet another component configuration, the particular outer component a may be adjacent to the inner component B and the inner component D, whereby damage to the particular outer component a may cause damage to the inner component B and the inner component D.
This damage map of the outer component to the inner component may be preconfigured. For example, the component structural configuration of the vehicle may be determined according to the model, the factory time, and the configuration level of the vehicle. For example, a mapping of the external component to one or more internal components may be established for each external component based on a component structural configuration to determine one or more internal components that are adjacent to the external component and thus may be damaged by damage to the external component.
This mapping of exterior components to interior components may also be determined based on empirical data relating to the proposed vehicle insurance (e.g., interior damaged components and exterior damaged components of the vehicle).
Vehicles may be classified according to their component configuration, and vehicles of each component configuration type may have the same component configuration, i.e., have the same set of damage maps of external components to one or more internal components.
Table 1 shows an example of a set of damage maps in a component structure configuration.
| |
Inner member A and inner member B |
| Outer part 2 | Inner member C and inner member D |
| …… | …… |
| Exterior part n | Inner member X, inner member Y, and inner member Z |
TABLE 1
Note that the number of internal components shown in table 1 is merely exemplary, one external component may map to more or fewer internal components, and two external components may also map to the same internal component.
The server 102 may include a lossy neural network model. The impairment Neural Network model may be a Convolutional Neural Network (CNN) or a regional recommendation Network (RPN).
The damage assessment neural network model may be constructed by training the model using the damage information for the car insurance committed. For example, the neural network model may be trained using a set of vehicle damage images of a vehicle on which the vehicle insurance has concluded, a component configuration type of the vehicle, an identification number of an external damaged component and its damage level, and an identification number of an internal damaged component.
If the damage condition of the vehicle to be damaged is to be determined, the server 102 may input the damage assessment image set and the component configuration type of the vehicle to be damaged into the trained damage assessment neural network model to determine the external damaged component and the damage degree thereof, and the internal damaged component and the damage probability thereof of the vehicle, so as to perform subsequent claim settlement operation.
The database 103 may store relevant information. For example, the related claim settlement information of the settled vehicle insurance includes a vehicle damage image set of the vehicle, a component configuration type of the vehicle, a vehicle external damaged component and a damaged degree, a vehicle internal damaged component and the like; the model of the vehicle, the time of delivery, the configuration level (e.g., high-mix, medium-mix, low-mix), the component configuration type; the number of the emergencies, the area, the vehicle price and the use property of the client in the previous year; vehicle component repair prices, etc.
Note that although server 102 and database 103 are shown separately in fig. 1, database 103 may be integrated with server 102.
Fig. 2 is a flow diagram of a method for image-based vehicle damage assessment according to aspects of the present disclosure. The method may be implemented by the server 102 of fig. 1.
At step 202, a set of impairment images of a vehicle may be acquired.
The server may obtain a set of impairment images for the vehicle. The mobile terminal can shoot a plurality of images on site and transmit the images to the server, and the third-party service platform can also obtain the images. The set of images of the damage to the vehicle may include multiple captured images of the vehicle or captured videos (a segment of a video may be considered a set of consecutive images). The vehicle damage image set may include a full vehicle image including license plate information, a distant view image of a vehicle damage location, a near view image of the vehicle damage location, and a vehicle damage detail image. If the user's vehicle has more than one damage, the set of vehicle damage images may include a distant view image of the vehicle damage site, a close view image of the vehicle damage site, and a vehicle damage detail image, which are respectively photographed for each damage, to completely record the entire vehicle damage information. The collection of impairment images may also include images of vehicle credentials, user credentials.
The server may filter the images in the set of vehicle damage images. For example, the quality of each image in the set of impairment images (e.g., sharpness, whether a vehicle impairment site was captured, etc.) may be determined. If the image quality is low or no lesion is captured, the image may be discarded and a message sent to the terminal asking the user to re-capture the image. The image quality determination can be processed by means of an ambiguity threshold, an information entropy value, and the like.
At step 204, relevant information for the vehicle may be obtained.
The identification code of the vehicle, the user identity information, and the like can be acquired.
The identification code of the vehicle may be obtained from the corresponding image in the set of loss images obtained in step 202. For example, the user may capture the identification code of the vehicle and include it in the set of loss images for transmission to the server. The server may recognize the vehicle identification code in the image through image processing. The vehicle identification code may also be entered by the user in the terminal and transmitted by the terminal to the server. The vehicle identification code may be used to determine a component configuration type of the vehicle.
The server can also obtain the user identity information, and further can determine the related information of the user according to the user identity information, such as the current annual insurance times of the user, the credit value of the user and the like.
At step 206, vehicle damage information may be determined from a set of damage images for the vehicle.
The vehicle damage information may include damage information of an external part and damage information of an internal part of the vehicle.
Vehicle impairment information may be determined using a trained impairment neural network model.
The damage assessment neural network model may be obtained by training the model using relevant information (a set of vehicle damage images, component configuration types, external component damage information, internal component damage information, etc.) of the vehicle that has been damaged. Wherein the external part damage information may include an identification number of the external damaged part and a damage degree thereof, and the internal part damage information may include an identification number of the internal damaged part.
In one aspect, a neural network model may be trained using a set of damage images of the determined damage vehicle risk, a component configuration type, internal component damage information, and external component damage information. Therefore, when the damage condition of the vehicle to be damaged is predicted, the input of the neural network model is a vehicle damage image set and a component configuration type, and the output is internal component damage information and external component damage information.
In another aspect, an external damage prediction neural network model and a plurality of internal damage prediction neural network models may be constructed, where each internal damage prediction neural network model corresponds to a vehicle component configuration type, i.e., has the same set of external component to internal component damage maps.
Fig. 3 is a diagram of an apparatus 300 for predicting vehicle damage according to an aspect of the present disclosure.
As shown in fig. 3, the apparatus 300 may include an external lesion prediction model 302, a plurality of internal lesion prediction modules 3041-N, and a selector 306.
The external damage prediction model 302 may determine external damaged component information of the vehicle, such as damaged external components and their damage types and damage levels, from a set of damage images of the vehicle.
The external damage prediction model 302 may be obtained by training the model using a set of damage images of the determined damage car risk and external damage information. The vehicle damage image set of the vehicle to be damaged may then be input into an external damage prediction model to determine (predict) external damage component information for the vehicle to be damaged.
The selector 306 may select the corresponding internal injury prediction model 304 according to the component configuration type of the vehicle.
Each of the internal injury prediction modules 3041-N corresponds to a component configuration type for a vehicle. The internal damage prediction model 304-i may be obtained by training the model using the damage image set and the internal damage information of the damaged vehicle having its component configuration type i. The vehicle damage image set of the vehicle to be damaged having component configuration type i may then be input into an internal damage prediction model to determine (predict) internal damaged component information, e.g., internal damaged components and damage probability, for the vehicle to be damaged.
The apparatus 300 for predicting vehicle damage of fig. 3 divides the prediction of internal damaged parts into a plurality of internal damaged part prediction models each corresponding to a different vehicle part configuration type, and can reduce the complexity of the neural network model of the apparatus 300, thereby reducing the training amount and the calculation amount of the neural network model.
Fig. 4 is a flow diagram of a method of training a vehicle damage assessment neural network model using samples of damaged vehicle insurance (a set of vehicle damage images, vehicle component configuration types, and vehicle component damage information), in accordance with aspects of the present disclosure.
The impairment neural network model may be trained using a large number of relevant sample data of the car insurance that has been impaired. These relevant sample data may include a set of vehicle damage images, component configuration types for the vehicle, determined external component damage information, and internal component damage information, such as information determined by a service professional through inspection of the vehicle.
At step 402, a set of damage images of vehicles at a determined damage risk may be obtained.
The server may retrieve from memory a set of impairment images for vehicles that have impaired vehicle insurance. The set of images of the car damage may be a set of images submitted and used by the user at the time of the car damage fix.
Additionally, the vehicle damage image set of the vehicle can be expanded, including rotating, translating, scaling, and color transforming the images in the image set, thereby improving the prediction accuracy of the model.
In step 404, damage information for a vehicle that has been determined to be at risk may be obtained.
The damage information may include damage information of the external part. Such as the identification number of the externally damaged component, and the degree of damage (e.g., mild, moderate, severe), etc.
The damage information may further include damage information for the internal component, such as an identification number of the internal damaged component, a degree of damage, and the like (e.g., mild, moderate, severe, and the like).
At step 406, a component configuration type of the vehicle may be determined.
Damage (e.g., denting, deformation) of exterior components of a vehicle may result in damage to interior components. A damage map may thus be established between the external component and one or more internal components that it may affect.
In one example, deformation of the front bumper skin may cause damage to the front bumper frame, the front bumper liner, and the radiator frame. A mapping can thus be established between the front bumper skin and the front bumper skeleton, front bumper liner, and radiator frame.
In another example, deformation of the hood may cause damage to the condenser, front bumper lining, radiator frame, hood hinges, and fan assembly. A mapping between the hood and condenser, front bumper lining, radiator frame, hood hinges, and fan assembly may thus be established.
Further, different vehicles may have different component configurations (placements). For example, the placement of various components in a vehicle may vary. Thus in different types of vehicles damage to the same exterior component may affect different interior components. Even for the same vehicle model, the placement positions of the respective components are often different due to different annual delivery times and different configuration levels (e.g., high-level, medium-level, and low-level).
The present disclosure may classify vehicles according to the placement locations of various components in the vehicles, each component configuration type of vehicle having the same component configuration, i.e., having the same set of mappings of outer components to one or more inner components. The server can determine the model, the factory time and the configuration level (high-configuration, medium-configuration and low-configuration) of the vehicle through the identification code of the vehicle, and further determine the component configuration type of the vehicle. The component configuration types of the vehicles of various models, factory hours, and configuration levels may be predetermined.
Generally, vehicles of the same model, time of manufacture, configuration level will have the same set of mappings. Vehicles of different models, time of delivery, configuration levels may also have the same set of mappings. The mapping set and thus the component configuration type of the vehicle can be determined according to the configuration structure of the vehicle.
At step 408, the neural network model may be trained using the set of impairment images for the determined impairment vehicle risk, the impairment information, and the component configuration type.
In the case of only one neural network model, a set of vehicle damage images of the determined damage vehicle risk, a vehicle component configuration type, damage information (external component damage information and internal component damage information) may be input to the neural network model for training. The external part damage information may include an external damaged part identification number and its damage level. The internal part damage information may include an internal damaged part identification number.
In an example of using the external damage prediction model and the plurality of internal damage prediction models to determine vehicle damage conditions shown in fig. 3, the external damage prediction model may be trained using the set of vehicle damage images and the external component damage information; and selecting a corresponding internal damage prediction model according to the vehicle component configuration type, and training the internal damage prediction model by using the vehicle damage image set, the external component damage information and the internal component damage information.
Returning to FIG. 2, at step 208, a damage assessment result may be determined based on the determined vehicle exterior component damage information and interior component damage information.
For damage to the external component, each degree of damage may be set to correspond to a maintenance solution. For example, severe deformation corresponds to part changes, mild deformation requires sheet metal, and mild scratches require painting. For a user, the final output of a damaged part can be a maintenance scheme, and when a damaged part has multiple damages, the final treatment mode of the whole part can be taken as the maintenance mode of the most seriously damaged part. Usually, one part of the vehicle is a whole, and if there are many damages, the most serious damage is reasonably treated. A maintenance scheme can be selected to solve all damages on a damaged part, for example, in one damaged part, the damage degree of one damaged part is serious damage and needs to be changed, the damage degree of the other damaged part is moderate deformation and needs to be made of sheet metal, and then the part can be selected to be changed without sheet metal processing.
For damage to the internal component, the determined probability of damage to the internal damaged component may be compared to a threshold. If the probability is above a threshold, a determination may be made that the internal injury component is to be claimed (e.g., replaced). If the probability is below a threshold, it may be determined that the internal injury component requires manual inspection to determine if there is injury.
Further, the server may obtain a repair cost for the component of the vehicle from the database to estimate a cost for the repair of the vehicle.
Specifically, the server may search for a repair price corresponding to the vehicle component in the repair strategy according to the damaged component, the damage degree, and the repair strategy of the vehicle component, and calculate an estimated repair price of the vehicle component.
For example, a different price library may be called according to information of a vehicle model to which the vehicle component belongs, a maintenance place of the selected vehicle component, a maintenance policy of a repair shop (for example, a 4S shop or a general integrated repair shop), and the like, and a maintenance plan including a pre-maintenance processing method and a corresponding estimated maintenance price for the vehicle component may be generated. The information of the maintenance strategy can be determined by the user's selection. For example, the user may select a repair location, whether at a 4S store or a complex repair shop, enter a vehicle make, model, and the algorithm may then determine a repair plan based on the repair strategy information for that vehicle component and the identified damaged component and degree of damage.
The server may send the determined damage information and the maintenance schedule back to the user for user query and confirmation.
Fig. 5 is a diagram of an apparatus for image-based vehicle damage assessment according to aspects of the present disclosure.
As shown in fig. 5, the apparatus 500 for image-based vehicle damage assessment may include an image acquisition module 502, an image processing module 504, a component configuration type determination module 506, an external damage determination module 508, an internal damage determination module 510, and a claims module 512.
The image acquisition module 502 may acquire a set of impairment images of the vehicle insurance. The set of impairment images may be obtained from a terminal of the user. The vehicle damage image set can comprise a whole vehicle image containing license plate information, a distant view image of a vehicle damage part, a close view image of the vehicle damage part and a vehicle damage detail image. The collection of impairment images may also include a photograph of a vehicle credential, a user credential.
The image processing module 504 may process images in the set of impairment images. For example, the damage site of the vehicle may be determined by recognition of a photograph of the vehicle; acquiring an identification code of a vehicle by processing an image of a vehicle certificate; the user identity information is obtained by processing the image of the user identity card, and the like.
Further, the quality of each image in the set of vehicle damage images (e.g., sharpness, whether a damage site is captured, etc.) may be determined, and the images in the set of vehicle damage images may be filtered according to the quality of the images.
The component configuration type determination module 506 may determine a component configuration type of the vehicle. The vehicle identification code may be used to determine the model number, the factory time, the configuration level, etc. of the vehicle, and thus the component configuration type of the vehicle. For example, the damage map set of the external component and the internal component may be determined according to the vehicle configuration structure diagram of each model of vehicle having the same factory time and configuration level, and the component configuration type may be determined accordingly. Vehicles having the same component configuration type have the same damage map of the external component and the internal component, and thus can contribute to determining the damage probability of the internal component by the damage condition of the external component.
The external damage determination module 508 may determine the damaged external component, its degree of damage, etc. from the set of impairment images.
The trained impairment neural network model may be used to determine the external components of the impairment, the extent of the impairment, etc. from the set of vehicle impairment images. The impairment Neural Network model may be a Convolutional Neural Network (CNN) or a regional recommendation Network (RPN).
The internal damage determination module 510 may determine internal damaged components of the vehicle and their damage probabilities from the set of vehicle damage images, the component configuration type of the vehicle.
Neural network models can be used to predict internal damaged components of a vehicle and their damage probabilities. Specifically, the neural network model may be trained using a set of vehicle damage images of a vehicle on which the vehicle insurance has concluded, a component configuration type of the vehicle, an external damaged component and a degree of damage thereof, and an identification number of an internal damaged component. The trained impairment neural network model may then be used to determine the interior damaged components of the vehicle and their damage probabilities from the set of vehicle impairment images, the component configuration types of the vehicle.
In one example, the external impairment determination module 508 and the internal impairment determination module 510 may be combined. For example, a neural network model may be trained using a set of damage images of the determined damage vehicle risk, component configuration types, internal component damage information, and external component damage information. Therefore, when the damage condition of the vehicle to be damaged is predicted, the input of the neural network model is a vehicle damage image set and a component configuration type, and the output is internal component damage information and external component damage information.
In another example, the external injury determination module 508 may include one external injury neural network model and the internal injury determination module 510 may include multiple internal injury neural network models, where each internal injury neural network model corresponds to one vehicle component configuration type. The external damage prediction model can determine damaged external parts of the vehicle and the damage degree of the damaged external parts according to the vehicle damage image set of the vehicle; each internal damage prediction model can determine internal damaged parts and damaged probability according to the vehicle damage image set, external damaged parts and damaged degree of the external damaged parts for the corresponding vehicle part configuration type.
The claim settlement module 512 can determine a repair plan and a claim amount based on the determined external damage information and internal damage information, in combination with other information.
For example, for damage to external components, each degree of damage may be set to correspond to a repair solution.
For internal damage, the determined probability of damage to the internally damaged component may be compared to a threshold. If the probability is above a threshold, it may be determined that a claim (e.g., a replacement) will be made for the internal damaged component. If the probability is below a threshold, it may be determined that the internal injury component requires manual inspection to determine if there is injury.
Note that, herein, the external component damage sample used for training the damage assessment neural network model includes the identification number of the external damage component and the damage degree thereof, the internal component damage sample may include the identification number of the internal damage component, and the output of the damage assessment neural network model includes the identification number of the external damage component and the damage degree thereof, and the identification number of the internal damage component and the probability thereof, but the disclosure is not limited thereto. The external and internal component damage samples used to train the impairment neural network model and the output of the impairment neural network model may include more or less information. For example, the internal component damage sample used to train the damage-assessment neural network model may also include the identification number of the internal damaged component and its damage level, and thus the output of the damage-assessment neural network model may include the identification number of the internal damaged component and its damage level.
According to the scheme, the damage condition of the external part and the damage condition of the internal part of the vehicle can be determined based on the vehicle damage image set provided by the user and the component configuration type of the vehicle and the constructed damage identification model, the maintenance scheme and the estimated claim settlement price of the vehicle are further automatically determined, manual intervention is not needed in the processing process, and the full-link automatic claim settlement scheme of the vehicle insurance is realized.
The illustrations set forth herein in connection with the figures describe example configurations and are not intended to represent all examples that may be implemented or fall within the scope of the claims. The term "exemplary" as used herein means "serving as an example, instance, or illustration," and does not mean "preferred" or "advantageous over other examples. The detailed description includes specific details to provide an understanding of the described technology. However, the techniques may be practiced without these specific details. In some instances, well-known structures and devices are shown in block diagram form in order to avoid obscuring the concepts of the described examples.
In the drawings, similar components or features may have the same reference numerals. Further, various components of the same component arrangement type may be distinguished by following the reference label by a dash and a second label that distinguishes among the similar components. If only the first reference label is used in the specification, the description is applicable to any one of the similar components having the same first reference label irrespective of the second reference label.
The various illustrative blocks and modules described in connection with the disclosure herein may be implemented or performed with a general purpose processor, a DSP, an ASIC, an FPGA or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices (e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration).
The functions described herein may be implemented in hardware, software executed by a processor, firmware, or any combination thereof. If implemented in software executed by a processor, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Other examples and implementations are within the scope of the disclosure and the following claims. For example, due to the nature of software, the functions described above may be implemented using software executed by a processor, hardware, firmware, hard-wired, or any combination thereof. Features that implement functions may also be physically located at various locations, including being distributed such that portions of functions are implemented at different physical locations. In addition, as used herein, including in the claims, "or" as used in a list of items (e.g., a list of items accompanied by a phrase such as "at least one of" or "one or more of") indicates an inclusive list, such that, for example, a list of at least one of A, B or C means a or B or C or AB or AC or BC or ABC (i.e., a and B and C). Also, as used herein, the phrase "based on" should not be read as referring to a closed condition set. For example, an exemplary step described as "based on condition a" may be based on both condition a and condition B without departing from the scope of the present disclosure. In other words, the phrase "based on," as used herein, should be interpreted in the same manner as the phrase "based, at least in part, on.
Computer-readable media includes both non-transitory computer storage media and communication media, including any medium that facilitates transfer of a computer program from one place to another. Non-transitory storage media may be any available media that can be accessed by a general purpose or special purpose computer. By way of example, and not limitation, non-transitory computer-readable media can comprise RAM, ROM, electrically erasable programmable read-only memory (EEPROM), Compact Disk (CD) ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other non-transitory medium that can be used to carry or store desired program code means in the form of instructions or data structures and that can be accessed by a general-purpose or special-purpose computer, or a general-purpose or special-purpose processor. Any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a web site, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL), or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. Disk (disk) and disc (disc), as used herein, includes CD, laser disc, optical disc, Digital Versatile Disc (DVD), floppy disk and blu-ray disc where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above are also included within the scope of computer-readable media.
The description herein is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to the disclosure will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other variations without departing from the scope of the disclosure. Thus, the disclosure is not intended to be limited to the examples and designs described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (17)
1. A vehicle damage assessment method comprising:
acquiring a vehicle damage image set of a vehicle;
determining a component configuration type for the vehicle, the component configuration type for the vehicle corresponding to a component configuration for the vehicle, the component configuration comprising a mapping of each exterior component of the vehicle to one or more interior components; and
inputting the component configuration type of the vehicle and the vehicle damage image set into a neural network model to determine internal damaged components of the vehicle and damage probabilities thereof.
2. The method of claim 1, further comprising:
processing the set of impairment images to determine an external impairment component of the vehicle; and
determining an internal damaged component of the vehicle and a damage probability thereof according to the component configuration type of the vehicle, the external damaged component and the vehicle damage image set.
3. The method of claim 1, further comprising:
processing the set of vehicle damage images to determine an external damaged component of the vehicle and its degree of damage; and
and determining a maintenance scheme of the vehicle according to the external damaged component and the damage degree thereof, and the internal damaged component and the damage probability thereof.
4. The method of claim 1, the component configuration type of the vehicle being determined by a vehicle identification code.
5. The method of claim 1, further comprising:
determining that the internal damaged component is to be claim if the probability of damage to the internal damaged component is above a threshold.
6. The method of claim 1, further comprising:
training the neural network model by using a training sample, wherein the training sample comprises a vehicle damage image set of the determined damage vehicle insurance, a component configuration type of the vehicle and an identification number of an internal damage component.
7. The method of claim 6, further comprising:
augmenting a set of images of the vehicle, the augmenting including rotating, translating, scaling, and color transforming images of the set of images.
8. The method of claim 1, further comprising:
determining a quality of each image in the set of impairment images; and
and screening the images in the vehicle damage image set according to the quality of the images.
9. A vehicle damage assessment device comprising:
means for obtaining a set of impairment images for a vehicle;
means for determining a component configuration type of the vehicle, the component configuration type of the vehicle corresponding to a component configuration of the vehicle, the component configuration comprising a mapping of each exterior component of the vehicle to one or more interior components; and
means for inputting the component configuration type of the vehicle and the set of vehicle damage images into a neural network model to determine an interior damaged component of the vehicle and its damage probability.
10. The apparatus of claim 9, further comprising:
means for processing the set of impairment images to determine an external impairment component of the vehicle; and
means for determining an interior damaged component of the vehicle and its damage probability from a component configuration type of the vehicle, the exterior damaged component, and the impairment image set.
11. The apparatus of claim 9, further comprising:
means for processing the set of impairment images to determine an external impairment component of the vehicle and its extent of impairment; and
means for determining a repair scenario for the vehicle based on the externally damaged component and its degree of damage, and the internally damaged component and its probability of damage.
12. The apparatus of claim 9, the component configuration type of the vehicle is determined by a vehicle identification code.
13. The apparatus of claim 9, further comprising:
means for determining that the internal damaged component is to be claimed if the probability of damage to the internal damaged component is above a threshold.
14. The apparatus of claim 9, further comprising:
a module for training the neural network model using training samples, the training samples including a set of impairment images of an impairment car risk, a component configuration type of a vehicle, an identification number of an internal impairment component.
15. The apparatus of claim 14, further comprising:
means for augmenting a set of images of the vehicle, the augmenting including rotating, translating, scaling, and color transforming images of the set of images.
16. The apparatus of claim 9, further comprising:
means for determining a quality of each image in the set of impairment images; and
and the module is used for screening the images in the vehicle damage image set according to the quality of the images.
17. A vehicle damage assessment device comprising:
a processor; and
a memory arranged to store computer executable instructions that, when executed, cause the processor to:
acquiring a vehicle damage image set of a vehicle;
determining a component configuration type for the vehicle, the component configuration type for the vehicle corresponding to a component configuration for the vehicle, the component configuration comprising a mapping of each exterior component of the vehicle to one or more interior components; and
inputting the component configuration type of the vehicle and the vehicle damage image set into a neural network model to determine internal damaged components of the vehicle and damage probabilities thereof.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010756372.2A CN111881856B (en) | 2020-07-31 | 2020-07-31 | Vehicle damage assessment method and device based on image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010756372.2A CN111881856B (en) | 2020-07-31 | 2020-07-31 | Vehicle damage assessment method and device based on image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111881856A true CN111881856A (en) | 2020-11-03 |
| CN111881856B CN111881856B (en) | 2023-01-31 |
Family
ID=73204329
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010756372.2A Active CN111881856B (en) | 2020-07-31 | 2020-07-31 | Vehicle damage assessment method and device based on image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111881856B (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112070250A (en) * | 2020-11-13 | 2020-12-11 | 深圳壹账通智能科技有限公司 | Vehicle damage assessment method and device, terminal equipment and storage medium |
| CN114241398A (en) * | 2022-02-23 | 2022-03-25 | 深圳壹账通科技服务有限公司 | Vehicle damage assessment method, device, equipment and storage medium based on artificial intelligence |
| US11403069B2 (en) | 2017-07-24 | 2022-08-02 | Tesla, Inc. | Accelerated mathematical engine |
| US11409692B2 (en) | 2017-07-24 | 2022-08-09 | Tesla, Inc. | Vector computational unit |
| US11487288B2 (en) | 2017-03-23 | 2022-11-01 | Tesla, Inc. | Data synthesis for autonomous control systems |
| US11537811B2 (en) | 2018-12-04 | 2022-12-27 | Tesla, Inc. | Enhanced object detection for autonomous vehicles based on field view |
| US11561791B2 (en) | 2018-02-01 | 2023-01-24 | Tesla, Inc. | Vector computational unit receiving data elements in parallel from a last row of a computational array |
| US11562231B2 (en) | 2018-09-03 | 2023-01-24 | Tesla, Inc. | Neural networks for embedded devices |
| US11567514B2 (en) | 2019-02-11 | 2023-01-31 | Tesla, Inc. | Autonomous and user controlled vehicle summon to a target |
| US11610117B2 (en) | 2018-12-27 | 2023-03-21 | Tesla, Inc. | System and method for adapting a neural network model on a hardware platform |
| US11636333B2 (en) | 2018-07-26 | 2023-04-25 | Tesla, Inc. | Optimizing neural network structures for embedded systems |
| US11665108B2 (en) | 2018-10-25 | 2023-05-30 | Tesla, Inc. | QoS manager for system on a chip communications |
| US11681649B2 (en) | 2017-07-24 | 2023-06-20 | Tesla, Inc. | Computational array microprocessor system using non-consecutive data formatting |
| US11734562B2 (en) | 2018-06-20 | 2023-08-22 | Tesla, Inc. | Data pipeline and deep learning system for autonomous driving |
| US11748620B2 (en) | 2019-02-01 | 2023-09-05 | Tesla, Inc. | Generating ground truth for machine learning from time series elements |
| US11790664B2 (en) | 2019-02-19 | 2023-10-17 | Tesla, Inc. | Estimating object properties using visual image data |
| US11816585B2 (en) | 2018-12-03 | 2023-11-14 | Tesla, Inc. | Machine learning models operating at different frequencies for autonomous vehicles |
| US11841434B2 (en) | 2018-07-20 | 2023-12-12 | Tesla, Inc. | Annotation cross-labeling for autonomous control systems |
| US11893393B2 (en) | 2017-07-24 | 2024-02-06 | Tesla, Inc. | Computational array microprocessor system with hardware arbiter managing memory requests |
| US11893774B2 (en) | 2018-10-11 | 2024-02-06 | Tesla, Inc. | Systems and methods for training machine models with augmented data |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170293894A1 (en) * | 2016-04-06 | 2017-10-12 | American International Group, Inc. | Automatic assessment of damage and repair costs in vehicles |
| CN110033011A (en) * | 2018-12-14 | 2019-07-19 | 阿里巴巴集团控股有限公司 | Traffic accident Accident Handling Method and device, electronic equipment |
| CN110569697A (en) * | 2018-08-31 | 2019-12-13 | 阿里巴巴集团控股有限公司 | Method, device and equipment for detecting components of vehicle |
| EP3629257A1 (en) * | 2018-09-28 | 2020-04-01 | Mitchell International, Inc. | Methods for estimating repair data utilizing artificial intelligence and devices thereof |
| CN111311540A (en) * | 2020-01-13 | 2020-06-19 | 平安科技（深圳）有限公司 | Vehicle damage assessment method and device, computer equipment and storage medium |
-
2020
- 2020-07-31 CN CN202010756372.2A patent/CN111881856B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170293894A1 (en) * | 2016-04-06 | 2017-10-12 | American International Group, Inc. | Automatic assessment of damage and repair costs in vehicles |
| CN110569697A (en) * | 2018-08-31 | 2019-12-13 | 阿里巴巴集团控股有限公司 | Method, device and equipment for detecting components of vehicle |
| EP3629257A1 (en) * | 2018-09-28 | 2020-04-01 | Mitchell International, Inc. | Methods for estimating repair data utilizing artificial intelligence and devices thereof |
| CN110033011A (en) * | 2018-12-14 | 2019-07-19 | 阿里巴巴集团控股有限公司 | Traffic accident Accident Handling Method and device, electronic equipment |
| CN111311540A (en) * | 2020-01-13 | 2020-06-19 | 平安科技（深圳）有限公司 | Vehicle damage assessment method and device, computer equipment and storage medium |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11487288B2 (en) | 2017-03-23 | 2022-11-01 | Tesla, Inc. | Data synthesis for autonomous control systems |
| US11409692B2 (en) | 2017-07-24 | 2022-08-09 | Tesla, Inc. | Vector computational unit |
| US11893393B2 (en) | 2017-07-24 | 2024-02-06 | Tesla, Inc. | Computational array microprocessor system with hardware arbiter managing memory requests |
| US11681649B2 (en) | 2017-07-24 | 2023-06-20 | Tesla, Inc. | Computational array microprocessor system using non-consecutive data formatting |
| US11403069B2 (en) | 2017-07-24 | 2022-08-02 | Tesla, Inc. | Accelerated mathematical engine |
| US11797304B2 (en) | 2018-02-01 | 2023-10-24 | Tesla, Inc. | Instruction set architecture for a vector computational unit |
| US11561791B2 (en) | 2018-02-01 | 2023-01-24 | Tesla, Inc. | Vector computational unit receiving data elements in parallel from a last row of a computational array |
| US11734562B2 (en) | 2018-06-20 | 2023-08-22 | Tesla, Inc. | Data pipeline and deep learning system for autonomous driving |
| US11841434B2 (en) | 2018-07-20 | 2023-12-12 | Tesla, Inc. | Annotation cross-labeling for autonomous control systems |
| US11636333B2 (en) | 2018-07-26 | 2023-04-25 | Tesla, Inc. | Optimizing neural network structures for embedded systems |
| US11562231B2 (en) | 2018-09-03 | 2023-01-24 | Tesla, Inc. | Neural networks for embedded devices |
| US11893774B2 (en) | 2018-10-11 | 2024-02-06 | Tesla, Inc. | Systems and methods for training machine models with augmented data |
| US11665108B2 (en) | 2018-10-25 | 2023-05-30 | Tesla, Inc. | QoS manager for system on a chip communications |
| US11816585B2 (en) | 2018-12-03 | 2023-11-14 | Tesla, Inc. | Machine learning models operating at different frequencies for autonomous vehicles |
| US11908171B2 (en) | 2018-12-04 | 2024-02-20 | Tesla, Inc. | Enhanced object detection for autonomous vehicles based on field view |
| US11537811B2 (en) | 2018-12-04 | 2022-12-27 | Tesla, Inc. | Enhanced object detection for autonomous vehicles based on field view |
| US11610117B2 (en) | 2018-12-27 | 2023-03-21 | Tesla, Inc. | System and method for adapting a neural network model on a hardware platform |
| US11748620B2 (en) | 2019-02-01 | 2023-09-05 | Tesla, Inc. | Generating ground truth for machine learning from time series elements |
| US11567514B2 (en) | 2019-02-11 | 2023-01-31 | Tesla, Inc. | Autonomous and user controlled vehicle summon to a target |
| US11790664B2 (en) | 2019-02-19 | 2023-10-17 | Tesla, Inc. | Estimating object properties using visual image data |
| CN112070250A (en) * | 2020-11-13 | 2020-12-11 | 深圳壹账通智能科技有限公司 | Vehicle damage assessment method and device, terminal equipment and storage medium |
| CN112070250B (en) * | 2020-11-13 | 2021-05-04 | 深圳壹账通智能科技有限公司 | Vehicle damage assessment method and device, terminal equipment and storage medium |
| WO2022100454A1 (en) * | 2020-11-13 | 2022-05-19 | 深圳壹账通智能科技有限公司 | Vehicle damage assessment method, apparatus, terminal device and storage medium |
| CN114241398A (en) * | 2022-02-23 | 2022-03-25 | 深圳壹账通科技服务有限公司 | Vehicle damage assessment method, device, equipment and storage medium based on artificial intelligence |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111881856B (en) | 2023-01-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111881856B (en) | Vehicle damage assessment method and device based on image | |
| US11823365B2 (en) | Automatic image based object damage assessment | |
| US11854181B2 (en) | Systems and methods for utilizing machine-assisted vehicle inspection to identify insurance buildup or fraud | |
| TWI710997B (en) | Auto insurance automatic compensation method and system | |
| US10430885B1 (en) | Processing insured items holistically with mobile damage assessment and claims processing | |
| US20130132166A1 (en) | Smart toll network for improving performance of vehicle identification systems | |
| CN112862702A (en) | Image enhancement method, device, equipment and storage medium | |
| US20210142464A1 (en) | System and method for artificial intelligence based determination of damage to physical structures | |
| US11735043B2 (en) | Systems and methods for fraud prevention based on video analytics | |
| US20220129860A1 (en) | Artificial Intelligence Based Determination Of Damage To Physical Structures Via Video | |
| CN111222409A (en) | Vehicle brand labeling method, device and system | |
| US20220311951A1 (en) | Systems and Methods for Gate-Based Vehicle Image Capture | |
| CN115984786A (en) | Vehicle damage detection method and device, terminal and storage medium | |
| CN111369373B (en) | Vehicle interior damage determination method and device | |
| CN113837170A (en) | Automatic auditing processing method, device and equipment for vehicle insurance claim settlement application | |
| DE102020135070A1 (en) | INFORMATION PROCUREMENT TECHNIQUES USING DATA FROM CONNECTED VEHICLES | |
| KR20200042779A (en) | System and method for recognizing and processing a fine bill automatically | |
| Kumar et al. | Parklot-Automated vehicle parking system using optical character recognition | |
| CN108876640A (en) | A kind of loss assessment system of image recognition and maintenance program estimation based on artificial intelligence | |
| KR102304921B1 (en) | Method for management of vehicle maintenance history by registering estimate | |
| CN113784009B (en) | Paper text image processing method and device and electronic equipment | |
| WO2021159634A1 (en) | Classification and regression dual-domain model based traffic identification method and system | |
| CN113486725A (en) | Intelligent vehicle damage assessment method and device, storage medium and electronic equipment | |
| CN113361424A (en) | Intelligent loss assessment image acquisition method, device, medium and electronic equipment for vehicle | |
| CN113807445A (en) | Method and device for file review, electronic device and computer program product |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20211220 Address after: Room 610, floor 6, No. 618, Wai Road, Huangpu District, Shanghai 200010 Applicant after: Ant Shengxin (Shanghai) Information Technology Co.,Ltd. Address before: 801-11, Section B, 8th floor, No. 556, Xixi Road, Xihu District, Hangzhou City, Zhejiang Province, 310023 Applicant before: Alipay (Hangzhou) Information Technology Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |