KR102109882B1 - System and method for diagnosing vehicle damage and calculating repair cost based on artificial intelligence - Google Patents

Abstract

According to an embodiment of the present invention, an artificial intelligence (AI) based vehicle damage diagnosis and repair cost estimation system and a method thereof. The AI based vehicle damage diagnosis and repair cost estimation system can comprise: a database that stores data related to estimation of a repair cost for a damage to a vehicle; a damage diagnosis part diagnosing the damage to the vehicle by analyzing an image of the vehicle input by a vehicle driver and determining damaged parts, damage types and the degree of the damage of the vehicle based on AI; a repair determination part determining a repair type for the damage based on a result of diagnosing the damage to the vehicle; and a repair cost estimation part estimating a repair cost for the damage based on data on the estimation of the repair cost stored in the database, the result of diagnosing and a result of determining the repair type.

Description

Artificial intelligence based vehicle damage diagnosis and repair cost calculation system and method {SYSTEM AND METHOD FOR DIAGNOSING VEHICLE DAMAGE AND CALCULATING REPAIR COST BASED ON ARTIFICIAL INTELLIGENCE}

The present invention relates to a vehicle damage diagnosis and repair cost calculation system and method for automatically diagnosing an accident vehicle damage and calculating a repair cost using artificial intelligence (AI).

As the spread of vehicles becomes more common, collisions of vehicles due to violations of traffic laws such as speeding and drunk driving are on the rise. Accordingly, the demand for repairing an accident vehicle is also gradually increasing.

However, it is a very cumbersome and time-consuming procedure for a driver to go to a maintenance company every time an accident occurs and request a diagnosis for a damaged part of the vehicle, and to receive a repair cost estimate.

In particular, as the technology level of vehicles has recently been advanced, it is difficult for the general public to recognize information on various parts and materials input to vehicles, and estimates of repair costs calculated by maintenance companies are also different. Therefore, it can be said that it is virtually impossible for the general public to judge whether the estimate of the repair cost calculated by the maintenance company is reasonable and reliable.

The present invention has been made to solve the above problems, and analyzes an image of an accident vehicle based on artificial intelligence to automatically diagnose a damaged portion and damage level of a vehicle, and calculate an appropriate repair cost estimate. It is an object of the present invention to provide a diagnosis and repair cost estimation system and method.

The AI-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention is a database that stores data related to calculation of a repair cost estimate due to vehicle damage, based on artificial intelligence (AI), and the vehicle input to the system Damage diagnosis unit for diagnosing the damage of the vehicle by determining the damaged parts, damage type, and degree of damage by analyzing the image of the vehicle, repair plate for determining the repair type for the damage based on the damage diagnosis result of the vehicle The government may include a repair cost calculation unit to calculate a repair cost according to the damage based on data related to the calculation of the repair cost estimate stored in the database, a diagnosis result of the damage, and a determination result of the repair type.

The AI-based vehicle damage diagnosis and repair cost calculation method according to an embodiment of the present invention is an AI-based vehicle damage diagnosis and repair cost calculation method using a system that includes a database that stores data regarding the calculation of repair cost estimates for vehicle damage, Based on AI, when an image of a vehicle is input to the system, diagnosing the damage by analyzing an image of the vehicle and determining a damaged part, a type of damage, and a degree of damage of the vehicle, based on a result of the damage diagnosis of the vehicle And determining a repair type for the damage, and calculating the repair cost according to the damage based on data related to the calculation of the repair cost estimate stored in the database, the diagnosis result of the damage, and the determination result of the repair type. can do.

According to the artificial intelligence-based vehicle damage diagnosis and repair cost calculation system and method according to the present invention, it is possible to automatically diagnose the damaged part and damage level of the accident vehicle and calculate an appropriate repair cost estimate, so that the driver of the vehicle can directly contact the maintenance company. Even if you don't visit, you can easily identify the damage history and get a rough estimate of the repair cost in advance.

1 is a block diagram of an artificial intelligence (AI) based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention.
2 shows an example of an automatic matching screen of damaged parts and repair details according to an embodiment of the present invention.
3 is a table illustrating an example of a criterion for classifying a damage class according to an embodiment of the present invention.
4A to 4C are images illustrating an example of a criterion for classifying damage classes according to FIG. 3.
5 shows an example of damage detection through artificial intelligence according to an embodiment of the present invention.
6 shows an example of a damage diagnosis and repair determination process according to an embodiment of the present invention.
7 shows an example of classifying the result of a repair decision according to an embodiment of the present invention.
8 shows an example of a damage diagnosis and repair method determination result according to an embodiment of the present invention.
9 shows an example of a result of calculating a repair cost according to a damage diagnosis and repair method determination result according to an embodiment of the present invention.
10 is a flowchart of an artificial intelligence based vehicle damage diagnosis and repair cost calculation method according to an embodiment of the present invention.
11 illustrates an example of artificial intelligence-based vehicle damage diagnosis and repair cost calculation system, a user (driver) terminal, and an insurance company server interlocking and transmitting data according to an embodiment of the present invention.
12 illustrates an example of linking insurance accident receipt information and automatic quotation calculation through recognition of a vehicle number in an image according to an embodiment of the present invention.
13A to 13D show examples of damage type diagnosis and damage depth determination for dent restoration repair according to an embodiment of the present invention.
14 is a flowchart illustrating that an artificial intelligence based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention operates with a user terminal.
15 is a hardware configuration diagram of an artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention.
16 is a block diagram of a user terminal for using an artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention.
17 is a hardware configuration diagram of a user terminal for using an artificial intelligence based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention.

Hereinafter, various embodiments of the present invention will be described in detail with reference to the accompanying drawings. In this document, the same reference numerals are used for the same components in the drawings, and duplicate descriptions for the same components are omitted.

With respect to various embodiments of the present invention disclosed in this document, specific structural or functional descriptions are exemplified only for the purpose of illustrating embodiments of the present invention, and various embodiments of the present invention may be implemented in various forms. And may not be construed as limited to the embodiments described in this document.

Expressions such as “first”, “second”, “first”, or “second” used in various embodiments may modify various components, regardless of order and / or importance, and It is not limited. For example, the first component may be referred to as a second component without departing from the scope of the present invention, and similarly, the second component may be referred to as a first component.

The terms used in this document are only used to describe specific embodiments, and may not be intended to limit the scope of other embodiments. Singular expressions may include plural expressions unless the context clearly indicates otherwise.

All terms used herein, including technical or scientific terms, may have the same meaning as commonly understood by a person skilled in the art of the present invention. Commonly used dictionary-defined terms may be interpreted as having the same or similar meaning to the meaning in the context of the related art, and should not be interpreted as an ideal or excessively formal meaning unless explicitly defined in this document. . In some cases, even terms defined in this document cannot be interpreted to exclude embodiments of the present invention.

1 is a block diagram of an artificial intelligence (AI) based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention.

Referring to FIG. 1, an artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention includes a database 110, a damage diagnosis unit 120, a repair determination unit 130, and a repair cost calculation unit 140 , The redundant image filtering unit 150 and the insurance information inquiry unit 160 may be included.

The database 110 may store data related to calculation of a repair cost due to damage to a vehicle. For example, the database 110 may store vehicle type information, vehicle parts list and price, vehicle damage type, labor, repair method and work time for each repair type, painting fee, and past repair cost estimate.

At this time, the vehicle type information may include the model name, options, grade, fuel efficiency, and the like of the vehicle, and the parts list of the vehicle may include a part name, a part number, and a price. In addition, the repair method for each repair type may include sheet metal, painting, repair, replacement, etc., and the labor information may include detachment time for each part, painting time, and labor cost according to the sheet metal time. In particular, in the case of the type of damage to the vehicle, the database 110 may be classified and stored according to damage classes according to established criteria such as damaged parts, damaged parts, and damage levels, as described later. However, this is only an example, and the information stored in the database 110 of the present invention is not limited to those listed.

In addition, the database 110 stores an image input to the system (for example, input by a user (vehicle driver) in the past) and uses it for a redundant image filtering function for preventing insurance fraud, as described below. Can be. In addition, the database 110 stores damaged images of a vehicle input by another user (driver), which can be used as a similar damage example to diagnose damage of the currently input vehicle image.

The damage diagnosis unit 120 analyzes an image of a vehicle input to the system (for example, input by a vehicle driver) based on artificial intelligence to determine a vehicle's damaged parts, damage type, and degree of damage to the vehicle. Can diagnose the damage. The damage diagnosis unit 120 may perform learning on the damage diagnosis of the vehicle using data related to the diagnosis of the vehicle damage previously processed in order to diagnose the damage of the artificial intelligence-based vehicle. In particular, the damage diagnosis unit 120 may classify the damage of the vehicle according to the damage class classified according to the damaged parts, the type of damage, and the degree of damage, as described later in FIG. 3. In this case, the damage class may be a result learned based on an existing damage diagnosis result processed by the damage diagnosis unit 120 based on AI. In addition, the damage diagnosis unit 120 recognizes the vehicle number of the image input by the vehicle driver, and as described below, can be used to interwork with the insurance accident acceptance information.

The repair determination unit 130 may determine the repair type for the damage based on the result of the damage diagnosis of the vehicle by the damage diagnosis unit 120 based on artificial intelligence. That is, the repair determination unit 130 may determine the type of parts and materials necessary for repair in the accident vehicle, a repair method, a work time, and the like, based on information stored in the database 110. In this case, the repair determination unit 130 may repeatedly learn data to determine each repair type corresponding to the damaged part, the type of damage, and the degree of damage to determine the artificial intelligence-based repair type.

In addition, the repair determination unit 130 may use a regression model when determining the repair type. Here, the regression model used in the repair determination unit 130 may be configured to diagnose damage using an equation derived to give different weights according to the damaged component, the type of damage, and the degree of damage. That is, the repair determination unit 130 may determine the repair type of the vehicle by applying a regression model to the damage result diagnosed by the damage diagnosis unit 120.

The repair cost calculation unit 140 may calculate the repair cost according to the damage based on artificial intelligence, based on data related to the calculation of the repair cost estimate stored in the database, a diagnosis result of the damage, and a determination result of the repair type. That is, the repair cost calculation unit 140 may calculate the final repair cost according to the damage to the vehicle by calculating work content and work time, material cost, part cost, and labor required for repair. At this time, the repair cost calculation unit 140 estimates the final repair cost through an arbitrary calculation formula based on data related to the calculation of the repair cost estimate stored in the database and the diagnosis result of the damage and the determination result of the repair type in order to calculate the repair cost based on artificial intelligence. Data can be trained repeatedly to calculate.

The duplicate image filtering unit 150 may detect whether an image input to the system (eg, input by a user) matches an image stored in the database 110. That is, the duplicate image filtering unit 150 has a function of automatically detecting whether a past billing history exists for each image input to automatically calculate an estimate based on artificial intelligence, and the vehicle driver intentionally Alternatively, it is possible to prevent claims for duplicate insurance due to negligence.

The vehicle driver inputs an image of the damaged vehicle for calculating and billing the repair cost, which can be transmitted to an insurance company, and in the process, an image of the past that has already been utilized by intention or carelessness can be uploaded. However, since it is often difficult for an insurer to receive an image to know whether to reuse the image, the image uploaded by the vehicle driver by the duplicate image filtering unit 150 is compared with the image already stored in the database 110 of the server to duplicate it. Whether or not it was automatically detected.

The insurance information inquiry unit 160 is interlocked with the insurance company's server, and can automatically query the vehicle's insurance reception information using the vehicle number. That is, when the damage diagnosis unit 120 automatically recognizes the vehicle number through the vehicle image input to the system, the insurance information inquiry unit 160 may automatically query the insurance company accident reception information with the corresponding vehicle number. Here, the accident reception information may include the type and duration of the insurance, the date and time of the accident, the vehicle number of the accident vehicle, the vehicle type, the insured person's name, the insured vehicle number, the insured vehicle type, the insurance money, etc. have.

2 shows an example of an automatic matching screen of damaged parts and repair details according to an embodiment of the present invention.

As shown in FIG. 2, when a vehicle driver photographs an image of an accident vehicle and inputs it to the artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention, the damaged part is automatically detected through the input image Then, the detected damaged parts are automatically matched to the labor costs and parts connected to the damaged parts in connection with the repair cost estimate.

Referring to FIG. 2, on the image of the accident vehicle, the entire damaged part area and the damaged part area may be displayed. In addition, for example, the price of each component may be displayed with respect to a damaged bumper cover, four-gram, absorber, bracket, beam, or the like. In addition, work division (main subject, painting, etc.), work content (contents of the work target), type of work (exchange, painting, desorption, sheet metal, etc.), work time, labor, part value, before ejaculation to repair damaged parts Details including parts, pre-assessment labor, reference parts, part numbers, etc. may also be displayed. For example, in the example of FIG. 2, the front bumper replacement operation (working time: 2.42 hours, labor cost: 72,600 won), the front bumper replacement (painting) (working time: 1.1890 hours, part cost 45,700 won, labor cost: 56,700 won ), Front bumper side bracket (left, right) (work time: 0.11 hours, labor cost: 3,300 won), headlamp (right) (work time: 0.38 hours, labor cost: 11,400 won), front suspension (right) ( Work time: 1.00 hours, labor amount: 30,000 won), advanced (right) sheet metal (work time: 2.780 hours, part amount 46,200 won, labor amount: 83,400 won), front door (right) (work time: 1.00 hours, Amount of wages: 30,000 won), front door (right) sheet metal (work time: 3.180 hours, part amount: 52,800 won, wage amount: 95,400 won), side mirror (electric, right) (work time: 0.38 hours, wage: 11,400 Won) is displayed.

However, this is only one example of an automatic matching screen of damaged parts and repair details, and the present invention is not limited to this, and a person skilled in the art may configure the screen in various ways that can be generally implemented.

3 is a table showing an example of a criterion for classifying a damage class according to an embodiment of the present invention.

Referring to FIG. 3, in order for artificial intelligence to recognize damage occurring on the exterior panel of an accident vehicle, the damage class is clearly classified and used to diagnose damage. At this time, the damage class classifies the damage of the vehicle according to the damaged parts, the type of damage, and the degree of damage. In the example of FIG. 3, the types of damage are classified into coating film damage, part detachment, material damage, and repairability, and these are classified into detailed items.

Specifically, as each detail item, coating film damage may include scratching and peeling damage. Part departure may include bumper (repairable) departure, bumper (replacement required) departure, fender departure, and door departure. Material damage can include scratch, surface damage, dent damage, linear damage, pressline damage, bend damage, pitting damage, complex damage. Non-repairable may include tearing, drilling, or crumpling. However, such details are exemplary and other details may be further included, or some of the details listed above may be excluded.

4A to 4C are images showing classification criteria in the damage class according to FIG. 3.

4A is a case where a front bumper of a vehicle is damaged, and only a coating film is damaged. In this case, the material under the coating film of the front bumper is intact. Therefore, the damage diagnosis unit 120 can determine the damage class of the vehicle as a coating film damage by grasping this (eg, based on the color of the damage site) through artificial intelligence.

4B is a case where the front bumper of the vehicle is damaged, and damage occurs even to the material under the coating film of the front bumper. Therefore, the damage diagnosis unit 120 can determine the damage class of the vehicle as material damage by grasping this (eg, based on scratches of the damage site) through artificial intelligence.

4C is a case in which the front bumper of the vehicle is damaged, and the damage to the front bumper is serious and repair is not possible. Therefore, the damage diagnosis unit 120 can determine the damage class of the vehicle as non-repairable by grasping this (eg, based on the degree of damage such as a crack in the damaged area) through artificial intelligence.

As shown in Figures 4a to 4c, the artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention, for example, damage to the coating film through the image of the damaged vehicle for the front bumper damage case It is possible to distinguish the damage class such as damage or non-repair.

In particular, the artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention is automatically damaged according to the damage class for any vehicle damage image based on artificial intelligence, that is, deep learning. It can be configured to diagnose. For example, by repeatedly learning the data so that the system can determine the damaged parts and the degree of damage through the image of the accident vehicle, if the user inputs only the image of the vehicle, the damaged part and depth of the damaged image are automatically performed without additional manipulation. Let the artificial intelligence determine your back.

5 shows an example of damage detection through artificial intelligence according to an embodiment of the present invention.

Referring to FIG. 5, a damaged component, a type of damage, and a degree of damage are detected according to the above-described damage class. As shown in FIG. 5, the artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention identifies a region in which a damage occurred in an input vehicle image (a square region in FIG. 5), and the damage occurs The parts (eg, 휀 / degree, bumper, etc.), the type and degree of damage (eg, material damage, coating film damage, repairability, etc.) can be displayed together.

6 shows an example of a damage diagnosis and repair determination process according to an embodiment of the present invention.

The repair determination unit according to an embodiment of the present invention may determine the final repair type of the damaged component using a regression model. That is, it is possible to determine the final repair type of the corresponding part by allocating the damaged area to the detected damage, calculating the parts and the damaged area, and determining the priority of the depth of damage among the generated damage.

Referring to FIG. 6, when damaged parts of a vehicle overlap each other, an area is divided based on an intersection and a positional relationship between two overlapped parts ((a) division of overlapping areas between parts). Next, the damages are matched to each part in consideration of the overlap ratio and range of damages to the part where the damage occurred (Fig. 6 (b) part-damage matching).

Then, the higher one of the depths of the damage in which the damage area is overlapped is selected (refer to (c) representative damage depth in FIG. 6). That is, when some of the multiple damage areas appearing on one component overlap, the damage with the highest degree of damage is selected. For example, when the scratched portion and the damaged portion overlap the bumper, the damaged portion having a high depth may be determined as a representative damage depth.

Then, the same damaged region adjacent to the same component among the damaged components is merged (the same component and the depth of damage of FIG. 6 are merged). For example, when a plurality of scratch areas and a plurality of break areas exist in the bumper where the damage has occurred, the scratch areas and the break areas can be merged.

For the merged damaged area, the ratio of the damaged area to the component area is calculated (Fig. 6 (e) damage area ratio calculation). At this time, for the entire area of each component, the ratio of the damaged area can be calculated by damage type or depth, and the ratio of the total damaged area to the total area of the component can also be calculated. Then, after calculating the repair time based on the calculated damage area ratio, at least the calculated repair time may be applied to the repair decision model to determine the repair type (FIG. 6 (f) repair type determination). At this time, the mathematical decision model may be the aforementioned regression model, but is not limited thereto, and various models and calculation formulas for determining the repair type may be applied.

7 shows an example of classifying the result of a repair decision according to an embodiment of the present invention.

Referring to FIG. 7, plastic parts (bumpers, etc.) and metal parts (fenders, doors, etc.) are classified at the time of repair determination, and the parts of each material are intact, scratched, repaired, replaced, or sheet metal ( Small / medium / large) classify repair types.

At this time, the above-described regression model may be applied to the repair determination, and the regression model may diagnose the damage using the equation derived to give different weights according to the damaged parts, the damage type, and the degree of damage, or determine the repair type. Can be configured.

For example, the following [Table 1] shows an example of a model for calculating the metal sheet metal working time.

Sheet metal working time calculation model for metal parts Classification Working time calculation formula fender Basic Sheet Metal Working Time (T _DS ) = 0.055 * (∑A _DC + ∑A _DP + ∑A _DN ) +0.6 Press Line Surcharge Time (T _DP ) = 0.0075 * ∑A _DP + 0.15 Bending / Chain Surcharge Time (T _DN ) = 0.02 * ∑A _DP + 0.15 door Basic Sheet Metal Working Time (T _DS ) = 0.05 * (∑A _DC + ∑A _DP + ∑A _DN ) +0.6 Press Line Surcharge Time (T _DP ) = 0.02 * ∑A _DP + 0.15 Bending / Chain Surcharge Time (T _DN ) = 0.04 * ∑A _DP + 0.15

Here, A _DC : surface damage, A _DP : press line damage, A _DN : bend and groove damage

At this time, the total sheet metal working time of the metal material component may be expressed as follows.

Total sheet metal working time (T _D ) of the part = basic sheet metal working time (T _DS ) + pressline premium time (T _DP ) + bending / dentation premium time (T _DN )

In addition, as described above, an example of a repair method according to a result of a sheet metal working time of a metal material component may be represented as the following [Table 2].

Repair method judgment model based on the result of sheet metal working time of metal parts Classification Determination of repair method
(T _D : Total sheet metal working time of the part) Fender Sheet metal (small) T _D ≤1.12h Sheet metal (medium) 0.88h <T _D ≤1.52h Sheet metal (large) 1.23h <T _D ≤2.07h exchange 1.73h <T _D door Sheet metal (small) T _D ≤1.45h Sheet metal (medium) 1.15h <T _D ≤1.85h Sheet metal (large) 1.5h <T _D ≤2.50h exchange 2.10h <T _D

That is, in the above-described example, when determining a repair method of a metal material component, an equation derived to give different weights according to a damaged component, a damage type, and the like is used with respect to a sheet metal working time. However, this is only an example, and the present invention is not limited thereto, and various calculation formulas in which appropriate weights are given according to damaged parts, damage types, and damage levels may be derived.

8 shows an example of a damage diagnosis and repair method determination result according to an embodiment of the present invention.

Referring to FIG. 8, it can be seen that detailed repair types and methods are determined according to each damage diagnosis performed according to artificial intelligence. For example, a repair method (eg, detachment, painting, replacement, etc. of each part) matched to each part may be determined for the final damage diagnosis result calculated by classifying the damage class as described above.

9 shows an example of a result of calculating a repair cost according to a damage diagnosis and repair method determination result according to an embodiment of the present invention.

Referring to FIG. 9, when the damaged vehicle image is analyzed based on artificial intelligence, damage to the “front fender (left)” among the parts of the vehicle occurs, and when repair judgment is determined to be “sheet metal (medium)” It shows the result of calculating the repair cost (damage to other parts is not shown). At this time, in the example of FIG. 9, the standard of the time for the repair and detachment of the insurance repair and the painting fee was 28,000 won per hour.

As shown in FIG. 9, the total repair cost estimate due to vehicle damage may be determined by adding labor, material cost, and part cost according to the working time. However, the present invention is not limited thereto, and the present invention is not limited to the calculation method illustrated in FIG. 9.

10 is a flowchart of an artificial intelligence based vehicle damage diagnosis and repair cost calculation method according to an embodiment of the present invention. The artificial intelligence-based vehicle damage diagnosis and repair cost calculation method of FIG. 10 may be an AI-based vehicle damage diagnosis and repair cost calculation method using a server including a database that stores data related to calculation of repair cost estimates according to vehicle damage.

Referring to FIG. 10, when an image of a vehicle is input to a system (for example, an application of a user terminal) by a user (vehicle driver), the image of the input vehicle is analyzed based on artificial intelligence, and damaged parts of the vehicle, The damage is diagnosed by determining the type and degree of damage (S110). For example, as described above, it is possible to specify a part at a vehicle damage site, and determine the type of damage such as coating film damage, part departure, material damage, and the like. In addition, the extent of damage can also be determined by calculating the area of the damaged area relative to the entire part area.

Then, the repair type for the damage is determined based on the result of the damage diagnosis of the vehicle (S120). Accordingly, as a result of the damage diagnosis, a repair method, a work time, parts and materials for repairing the damage may be calculated.

Finally, the repair cost according to the damage is calculated based on the data related to the calculation of the repair cost estimate stored in the database, the diagnosis result of the damage, and the repair type determination result (S130). As described above, it is possible to estimate an appropriate repair cost estimate by summing labor, material cost, and part cost according to the working time used to repair the damage.

As described above, according to the artificial intelligence-based vehicle damage diagnosis and repair cost calculation system and method according to an embodiment of the present invention, even if the driver of the accident vehicle directly visits a maintenance company and does not inquire for a quote, only inputs an image of the damaged vehicle In addition, it is possible to diagnose damage and calculate an approximate repair cost estimate, thereby reducing the hassle and securing reliability in estimating the repair cost.

11 illustrates an example of artificial intelligence-based vehicle damage diagnosis and repair cost calculation system, a user (driver) terminal, and an insurance company server interlocking and transmitting data according to an embodiment of the present invention.

Here, the server 100 may include a database and perform an artificial intelligence-based vehicle damage diagnosis and repair cost calculation function according to an embodiment of the present invention.

In addition, the user terminal 200 may include a tablet, a PC, a mobile phone, and the like, and communicate with the server 100 through an application that performs an artificial intelligence-based vehicle damage diagnosis and repair cost calculation function to transmit and receive data. .

That is, when a user (vehicle driver) photographs a damaged image of the vehicle through the user terminal 200 and transmits it to the server 100, for example, the server 100 uses the artificial intelligence to generate the vehicle image. And analyze the damage. Next, the repair type for the damage is determined based on the damage diagnosis result of the vehicle, and finally, based on the data on calculating the repair cost estimate stored in the database of the server 100, the diagnosis result of the damage, and the judgment result of the repair type Estimates of repair costs due to damage can be estimated.

However, this is only one example, and instead of the server 100, the application installed in the user terminal 200 analyzes the vehicle image itself to perform damage diagnosis, determines the repair type for the damage, and calculates the repair cost estimate It may be configured to.

On the other hand, the server 100 may communicate with the insurance company server 300, and as described above, may use the vehicle number detected as the vehicle image to automatically query the vehicle's insurance acceptance information. In addition, as well as the server 100, the user terminal 200 directly transmits the vehicle image photographed with the vehicle number to the insurance company server 300, so that the insurance company server 300 uses the vehicle number of the vehicle image itself. It may be configured to automatically inquire insurance receipt information and transmit it to the server 100 or the user terminal 200.

12 shows an example of linking insurance accident receipt information and automatic estimation of calculation through vehicle number recognition in an image according to an embodiment of the present invention.

The insurance accident acceptance information linkage function according to an embodiment of the present invention automatically recognizes the vehicle number using the license plate image of the photographed vehicle and provides details of the insurance receipt through linkage with the insurance company server. At this time, after recognizing the vehicle number from the vehicle license plate image based on the OCR technology, it is possible to automatically search the insurance company's accident information. In addition, detailed models and trims of the vehicle can be inquired based on the inquired accident reception information.

Particularly, the above-described function may be executed on the mobile application of the user terminal, and an example of an application for automatically recognizing the vehicle license plate and interlocking insurance information with the automatic calculation of the repair cost estimate of the damaged vehicle described above in FIG. 12 is shown.

Referring to FIG. 12, a) the main screen of the mobile application, b) vehicle license plate shooting and recognition function, c) vehicle number-insurance reception information server interlocking function, d) damaged part shooting function, e) damaged part determination function, f ) Represents an example of a repair method and a result of calculating a repair cost, which are only examples, and the system (or application) according to the present invention is not limited thereto.

13A to 13D show examples of damage type diagnosis and damage depth determination for dent restoration repair according to an embodiment of the present invention.

According to an artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention, artificial intelligence analysis is performed on a dent scope photographed image of a damaged area, including a function of calculating a repair cost for minor damage to a vehicle body After applying the technique and calculating the repair cost of the dent restoration, the calculation details can be presented on the screen.

Accordingly, it is possible to prevent an unfair charge of sheet metal and painting costs for minor damage that is not necessary for sheet metal and painting work by a maintenance company, and damage through a dent scope image analysis of minor damage areas that cannot be identified by a general image photograph. Diagnose types and estimate repair costs.

As described above, in the examples of FIGS. 13A to 13D, a dent scope photographed image may be additionally included in an image of a vehicle input to the system (eg, input by a user), and the damage diagnosis unit 120 of FIG. 1 The dent scope image is analyzed to determine the damage type and degree of damage of the vehicle's dent, and the repair cost calculation unit 140 may calculate the restoration repair cost of the dent based on the type and degree of damage of the dent.

13A to 13D, FIGS. 13A and 13B show types of dent damage (type V /-) and dent scope images therefor, and FIGS. 13C and 13D show damage using dent scope images for each type of dent. An example of depth determination is shown.

14 is a flowchart illustrating that an artificial intelligence based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention operates with a user terminal.

Referring to FIG. 14, first, a user (vehicle driver) photographs a damaged image of an accident vehicle and transmits it to the server (S210). The server analyzes the received image of the vehicle to diagnose damage to the vehicle (S220). Since the detailed damage diagnosis method is the same as that described with reference to FIGS. 1 to 13 above, a detailed description is omitted.

Next, the repair type for the damage is determined based on the damage diagnosis result of the vehicle (S230). Then, the repair cost according to the damage is calculated based on the data on the calculation of the repair cost estimate stored in the database of the artificial intelligence-based vehicle damage diagnosis and repair cost estimation system, the diagnosis result of the damage, and the repair type determination result (S240). The series of processes of S220 to S240 may be performed based on artificial intelligence embedded in the system as described above.

Finally, the repair cost estimate calculated by the server is transmitted to the user terminal (S250). At this time, the repair cost estimate received from the user terminal may be displayed on the display to the user.

In FIG. 14, it is described that the server calculates and estimates a repair cost according to vehicle damage, and transmits it to a user terminal. However, the present invention is not limited thereto, and the S220 to S240 are applied on an application for artificial intelligence based vehicle damage diagnosis and repair cost estimation of the user terminal. It may be possible to ensure that the process is performed.

According to the AI-based vehicle damage diagnosis and repair cost calculation system and method according to the present invention, it is possible to automatically diagnose the damaged part and damage level of the accident vehicle and calculate an appropriate repair cost estimate, so that the driver of the vehicle does not visit the maintenance company. Even if it is not, the damage history can be easily identified and the approximate repair cost estimate can be recognized in advance.

15 is a hardware configuration diagram of an artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention.

15, the artificial intelligence-based vehicle damage diagnosis and repair cost calculation system 100 according to an embodiment of the present invention is a CPU or GPU (hereinafter referred to as "CPU") 101, memory 102, It may include a communication interface 103.

The CPU 101 processes various data of the artificial intelligence-based vehicle damage diagnosis and repair cost calculation system 100 according to an embodiment of the present invention, and may be a processor that performs the functions described above. That is, the CPU 101 analyzes the image of the input vehicle based on the input from the user (for example, the vehicle image) transmitted from the user through the communication interface 103 to be described later by the system 100. It is possible to diagnose the vehicle damage by determining the damaged parts, the type of damage, and the degree of damage. Here, the interface input by the user will be described in detail by an application described through FIGS. 16 to 18 and a user terminal executing the application. In addition, the CPU 101 can cause the system 100 to determine the type of repair for damage based on the result of the vehicle damage diagnosis, and the data on the calculation of the repair cost estimate stored in the memory 102 and the diagnosis result of the damage And a repair cost according to damage based on a result of the repair type determination. In addition, the CPU 101 may also perform functions corresponding to the redundant image filtering unit and insurance information inquiry unit described in FIG. 1.

The memory 102 may store data related to calculation of a repair cost due to damage to a vehicle. That is, the memory 102 may be an example of the database 110 described above. For example, as described above, the memory 102 stores vehicle type information, vehicle parts list and price, vehicle damage type, labor, repair method and repair time for each repair type, painting fee, and historical repair cost estimate history. Can be. In addition, the memory 102 may store an image input by the user in the past for a duplicate image filtering function to prevent insurance fraud.

 A plurality of such memory 102 may be provided as needed. The memory 102 may be a volatile memory or a nonvolatile memory. As the volatile memory, RAM, DRAM, SRAM, or the like may be used. As the memory 102 as a non-volatile memory, ROM, PROM, EAROM, EPROM, EEPROM, flash memory and the like can be used. The examples of the memory 102 listed above are merely examples and are not limited to these examples.

In addition, the memory 102 may store programs, such as software and firmware, that control the operation of the vehicle damage diagnosis and repair cost calculation system and perform functions. By executing these programs by the CPU 101, each function configuration of the vehicle damage diagnosis and repair cost calculation system, namely, the damage diagnosis unit 120, the repair determination unit 130, the repair cost calculation unit 140, and the redundant image filtering unit ( 150), the insurance information inquiry unit 160 may be implemented.

The communication interface 103 communicates with external devices such as the user terminal 200 and the insurance company server 300 to exchange various data or signals. The communication interface 103 may be configured to communicate with an external device wirelessly or wired. The above-described vehicle image or the like may be received from the user terminal 200 through the communication interface 103. In addition, information on damaged parts, types of damage, and degree of damage of the vehicle as a result of analysis based on the received vehicle image may be transmitted to the user terminal 200 through the communication interface 103 again. Similarly, a signal of an inquiry request regarding insurance acceptance information may be transmitted to the insurance company server 300 through the communication interface 103, and the result may be received again from the insurance company server 300.

Although not shown, an input / output interface for interoperating with an operator or manager of the artificial intelligence based vehicle damage diagnosis and repair cost calculation system 100 according to an embodiment of the present invention may be further included.

As an example of the input / output interface, for example, an input device such as a keyboard, a mouse, or a touch panel may be included, and a user may perform an operation for inputting an image of a damaged vehicle. The input device generates an operation signal based on the operation from the user as described above and transmits it to the CPU 101 or other corresponding hardware so that the operation according to the operation signal is expressed.

As an example of the input / output interface, for example, an output device such as a display may be included. The display may display various screens generated by the operation of the artificial intelligence based vehicle damage diagnosis and repair cost calculation system 100 to an operator or a manager. In addition, a user interface (UI) may be displayed to an operator or an administrator through a display.

16 is a block diagram of a user terminal for using an artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention.

Referring to FIG. 16, the user terminal 200 may include a memory unit 210, a communication unit 220, an input unit 230, a control unit 240, and a display unit 250.

The memory unit 210 may store applications for artificial intelligence-based vehicle damage diagnosis and repair cost calculation. The application may be created and distributed by a subject operating an artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention, or may be written by a third party. Such an application may include a communication module, a damage diagnosis module, a repair cost calculation module, etc. that can be implemented by being executed by the control unit 240 described later.

Also, the memory unit 210 may store data related to diagnosis of vehicle damage and calculation of a repair cost estimate due to vehicle damage.

As described above, the communication unit 220 may function to communicate with a server by executing a communication module by executing an application. The communication unit 220 may transmit a photo of a damaged vehicle to a server, a request for an estimate of the repair cost, and various information or data necessary for this. In addition, the communication unit 220 analyzes the diagnosis results based on various data transmitted from the server.

The input unit 230 may allow a user to input or select a desired function, such as a keyboard or a touch panel. For example, in an embodiment of the present invention, a user may input an image of a damaged vehicle through the input unit 230. Alternatively, the input unit 23 may be a device such as a camera for a user to directly take an image of a damaged vehicle.

The controller 240 may perform various functions for diagnosing damage to a vehicle and estimating a repair cost by executing an application stored in the memory 210. That is, when the user executes the application and requests a repair cost estimate based on the photograph of the damaged vehicle, the controller 240 transmits the photograph of the damaged vehicle and various data accompanying it to the server. In addition, the control unit 240 receives the diagnostic information about the damaged parts, damage types, damage degree, repair cost, etc., which is an analysis result through the communication unit 220.

Alternatively, the application installed in the user terminal 200 may perform damage diagnosis and repair cost estimation of the vehicle itself. In this case, the above operation may be implemented by executing a damage diagnosis module, a repair cost calculation module, and the like included in the application. That is, when the user inputs the vehicle image through the input unit 230, the control unit 240 analyzes the image of the vehicle input by the vehicle driver as described above to determine the damaged parts, damage type, and degree of damage of the vehicle. Vehicle damage can be diagnosed. In addition, the control unit 240 determines the repair type for the damage based on the damage diagnosis result of the vehicle, and calculates the repair cost according to the damage based on the data on calculating the repair cost estimate and the diagnosis result of the damage and the determination result of the repair type You can make it perform all functions.

The display unit 250 performs an operation according to an artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention, such as an image of a damaged vehicle, a damage diagnosis history, a repair type determination history, and a repair cost calculation result through an application. Can produce one result.

17 is a hardware configuration diagram of a user terminal for using an artificial intelligence-based vehicle damage diagnosis and repair cost calculation system according to an embodiment of the present invention.

Referring to FIG. 17, the user terminal 200 for diagnosing vehicle damage based on artificial intelligence and calculating repair cost according to an embodiment of the present invention includes a CPU or GPU (hereinafter referred to as “CPU”) 201, memory ( 202), a user interface 203, a display 204, and a speaker 205.

The CPU 201 executes the application 400 stored in the memory 202 for diagnosing and repairing vehicle damage based on artificial intelligence according to an embodiment of the present invention, and artificial intelligence according to an embodiment of the present invention It may be a processor that processes various data for diagnosing vehicle damage and calculating repair costs, and performs the functions described above. That is, the CPU 201 may perform the functions of the damage diagnosis unit, the repair determination unit, the repair cost calculation unit, the redundant image filtering unit, and the insurance information inquiry unit by executing the application 300 stored in the memory 202. Since the function of the CPU 201 is substantially the same as that of the CPU 101 described in FIG. 15, detailed description thereof will be omitted.

The memory 202 may store an application 400 for artificial intelligence-based vehicle damage diagnosis and repair cost estimation, and may also store various data related to vehicle damage diagnosis and estimate of repair cost according to damage. Since the memory 202 is substantially the same as the memory 102 described with reference to FIG. 15, a detailed description is omitted.

The input interface 203 may be, for example, an input device such as a keyboard, mouse, or touch panel, and allows a user to perform an operation for inputting an image of a damaged vehicle.

The display 204 is the result of performing the operation of the artificial intelligence-based vehicle damage diagnosis and repair cost calculation application 400 according to an embodiment of the present invention, such as an image of a damaged vehicle, a damage diagnosis history, a repair type determination history, and a repair cost calculation result. Can represent The speaker 205 may transmit a voice to the user regarding the content displayed on the display 204.

The communication interface 206 is a configuration capable of transmitting and receiving various data to and from the server. The communication interface 206 may be various devices capable of supporting wired or wireless communication.

In the above, even if all the components constituting the embodiment of the present invention are described as being combined or operated as one, the present invention is not necessarily limited to these embodiments. That is, if it is within the scope of the present invention, all of the components may be selectively combined and operated.

In addition, the terms "include", "consist" or "have" as described above mean that the corresponding component can be intrinsic, unless specifically stated otherwise, to exclude other components. It should not be interpreted as being able to further include other components. All terms, including technical or scientific terms, have the same meaning as generally understood by a person skilled in the art to which the present invention belongs, unless otherwise defined. Commonly used terms, such as predefined terms, should be interpreted as being consistent with the contextual meaning of the related art, and are not to be interpreted as ideal or excessively formal meanings unless explicitly defined in the present invention.

The above description is merely illustrative of the technical idea of the present invention, and those of ordinary skill in the art to which the present invention pertains may make various modifications and variations without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical spirit of the present invention, but to explain, and the scope of the technical spirit of the present invention is not limited by these embodiments. The scope of protection of the present invention should be interpreted by the claims below, and all technical spirits within the scope equivalent thereto should be interpreted as being included in the scope of the present invention.

100: vehicle damage diagnosis and repair cost calculation system
110: database 120: damage diagnosis unit
130: repair judgment unit 140: repair cost calculation
150: duplicate image filtering unit 160: insurance information inquiry unit
200: user terminal 210: memory unit
220: communication unit 230: input unit
240: control unit 250: display unit
300: insurance company server 400: application
410: storage module 420: communication module
430: damage diagnosis module 440: repair cost calculation module

Claims (20)

AI-based vehicle damage diagnosis and repair cost calculation system,
A database for storing data related to estimating repair cost due to damage to a vehicle and images input in the past into the system;
A damage diagnosis unit for diagnosing damage to the vehicle based on artificial intelligence (AI), by analyzing an image of the vehicle input to the system and determining a damaged part, damage type, and degree of damage of the vehicle;
A repair determination unit determining a repair type for the damage based on a result of the damage diagnosis of the vehicle;
A repair cost calculation unit for calculating a repair cost according to the damage based on data related to the calculation of the repair cost estimate stored in the database, the diagnosis result of the damage, and the determination result of the repair type; And
And a duplicate image filtering unit that compares an image input to the system with an image input in the past stored in the database to detect whether there is a past billing history for the image input to the system.
The damage diagnosis unit,
When the damaged parts of the vehicle overlap each other, an area is divided based on an intersection and a positional relationship of the overlapped parts on the image,
The damage is matched for each part based on the damage overlap ratio to the damaged part of the vehicle,
When the damaged area overlaps, the area with the highest degree of damage among the overlapped damaged areas is determined as a representative area,
Merging damage zones belonging to the same damage type adjacent to the same part,
The repair judgment unit,
AI based vehicle damage diagnosis and repair cost calculation system to determine the repair type of the damage by calculating the ratio of the damaged area merged to the total area of the part. The method according to claim 1,
The database includes AI-based vehicle damage diagnosis and repair cost calculation including the vehicle type, parts list and price of the vehicle, vehicle damage type, labor cost, repair method and work time for each repair type, painting fee, and past repair cost estimates. system. The method according to claim 1,
The damage diagnosis unit classifies the damage of the vehicle according to the damage class classified according to the damaged parts, the damage type, and the degree of damage, wherein the damage class is based on the existing damage diagnosis result processed by AI based on the damage class. AI-based vehicle damage diagnosis and repair cost estimation system, characterized by the learned results. The method according to claim 1,
The repair determination unit AI-based vehicle damage diagnosis and repair cost calculation system to determine the repair type of the vehicle by applying a regression model to the damage result diagnosed by the damage diagnosis unit. The method according to claim 4,
The regression model is an AI-based vehicle damage diagnosis and repair cost calculation system using equations derived to give different weights according to the damaged parts, the type of damage, and the degree of damage. The method according to claim 1,
The repair determination unit calculates a ratio of the damaged area matched to the total area of the part, and calculates a repair time according to the calculated ratio, thereby determining an repair type of the damage based on AI-based vehicle damage diagnosis and repair cost calculation system . delete delete The method according to claim 1,
The damage diagnosis unit recognizes the vehicle number of the image input to the system,
AI-based vehicle damage diagnosis and repair cost calculation system that is interlocked with the insurance company's server, and further includes an insurance information inquiry unit for automatically inquiring the insurance receipt information of the vehicle using the vehicle number. The method according to claim 1,
The image of the vehicle input to the system includes a dent scope shooting image,
The damage diagnosis unit analyzes the dent scope image to determine the type and degree of damage to the dent of the vehicle,
The repair cost calculation unit is based on the damage type and the degree of damage of the dent, the AI-based vehicle damage diagnosis and repair cost calculation system for calculating the repair repair cost of the dent. As an AI-based vehicle damage diagnosis and repair cost calculation method using a system that includes a database for storing estimates of repair costs due to vehicle damage and a database storing images input in the past,
Based on AI, when an image of a vehicle is input to the system, diagnosing the damage by analyzing an image of the vehicle and determining a damaged part, a type of damage, and a degree of damage of the vehicle;
Determining a repair type for the damage based on a result of the damage diagnosis of the vehicle;
Calculating a repair cost according to the damage based on data related to the calculation of the repair cost estimate stored in the database, a diagnosis result of the damage, and a determination result of the repair type; And
Comparing the image input to the system and the image input in the past stored in the database to detect whether there is a past billing history for the image input to the system, and
Diagnosing the damage,
If the damaged parts of the vehicle overlap each other, dividing an area based on a positional relationship and an intersection of the overlapped parts on the image;
Matching damage for each of the parts based on a damage overlap ratio to the damaged parts of the vehicle;
If the damaged area is overlapped, determining an area having the highest degree of damage among the overlapped damaged areas as a representative area; And
And merging damage regions belonging to the same damage type adjacent to the same component,
The step of determining the repair type,
And determining a repair type of the damage by calculating a ratio of the damaged area merged to the total area of the component. The method according to claim 11,
The database includes AI-based vehicle damage diagnosis and repair cost calculation including the vehicle type, parts list and price of the vehicle, vehicle damage type, labor cost, repair method and work time for each repair type, painting fee, and past repair cost estimates. Way. The method according to claim 11,
The step of diagnosing the damage may include classifying the damage of the vehicle according to the damage class classified according to the damaged parts, the type of damage, and the degree of damage. The method according to claim 11,
The determining of the repair type includes determining a repair type of the vehicle by applying a regression model to the diagnosed damage result, and the AI-based vehicle damage diagnosis and repair cost calculation method. The method according to claim 14,
The regression model is an AI-based vehicle damage diagnosis and repair cost calculation method using equations derived to give different weights according to the damage component, the damage type, and the degree of damage. The method according to claim 11,
The step of determining the repair type includes determining a repair type of the damage by calculating a ratio of the damaged area matched to the total area of the part and calculating a repair time according to the calculated rate. Based vehicle damage diagnosis and repair cost calculation method. delete delete The method according to claim 11,
Diagnosing the damage includes recognizing the vehicle number of the image input to the system,
AI-based vehicle damage diagnosis and repair cost calculation method further comprising the step of automatically inquiring the insurance acceptance information of the vehicle using the vehicle number in conjunction with an insurance company server. The method according to claim 11,
The image of the vehicle input to the system includes a dent scope shooting image,
The step of diagnosing the damage includes determining the type and degree of damage of the dent of the vehicle by analyzing the dent scope image.
The calculating of the repair cost according to the damage includes calculating the repair repair cost of the dent based on the damage type and the degree of damage of the dent, and an AI-based vehicle damage diagnosis and repair cost calculation method.

Images