VISUAL ANALYTICS BASED VEHICLE INSURANCE ANTI-FRAUD DETECTION

• US 20180240194A1
• Filed: 02/23/2017
• Published: 08/23/2018
• Est. Priority Date: 02/23/2017
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: DONG, Wei Shan, GAO, Peng, LI, Chang Sheng, LUO, Wen Han, YAO, Ren Jie, YUAN, Ting, ZHU, Jun
• Granted Patent: US 10,628,890 B2
• Time in Patent Office: Days
• CPC Class Codes: G06Q 30/0185 Product, service or business identity fraud
• CPC Class Codes: G06Q 40/08 Insurance

Abstract

A system, a computer readable storage medium, and a method for detecting insurance fraud or for comparing references use visual analytics-based techniques. The method can include identifying a scratch and a scratch location on a vehicle in a 3-dimensional rendering in comparison to a base image or in comparison to a stored image in a database, comparing one or more features of the scratch such as color features of the scratch in the scratch location in comparison to the scratch location or comparing one or more texture features of the scratch in the scratch location in comparison to the scratch location in the base image or the stored image. The method can further generate a similarity calculation based on the one or more comparisons for the scratch, the scratch location, the one or more scratch features of the scratch and presents a result in response to the similarity calculation.

Claims

1. A computer implemented method, comprising:

identifying, by at least one processor of at least one computing device, a scratch and a scratch location on a vehicle in a 3-dimensional rendering in comparison to a base image;

comparing, by the at least one or more processors, one or more color features of the scratch in the scratch location in comparison to the scratch location in the base image;

comparing, by the at least one or more processors, one or more texture features of the scratch in the scratch location in comparison to the scratch location in the base image;

generating, by the at least one or more processors, a similarity calculation based on one or more comparisons for the scratch, the scratch location, the one or more color features of the scratch, and the one or more texture features of the scratch; and

presenting a result in response to the similarity calculation.

2. The computer implemented method of claim 1, wherein the method further comprises confirming an insurance claim based on the similarity calculation having a score above a predetermined threshold.

3. The computer implemented method of claim 1, wherein the identifying comprises a vehicle pose estimation, localization of the scratch location to a scratch area and a mapping to a vehicle surface location by unfolding a 3-dimensional vehicle model.

4. The computer implemented method of claim 1, wherein the comparing of one or more texture features comprises an extraction of texture features using at least one of deep learning techniques, deep neural networks, or extreme learning machines.

5. The computer implemented method of claim 1, wherein the comparing of one or more color features comprises use of a hue, saturation, and value histogram.

6. The computer implemented method of claim 1, wherein the similarity calculation comprises an analysis of the comparison of the scratch location and a comparison of metadata associated with the base image or with a stored image in a database.

7. The computer implemented method of claim 1, further comprising comparing, by the at least one or more processors, a plurality of images for consistency for vehicle model, scratch location, geographic location tag, and time tag.

8. The computer implemented method of claim 7, further comprising comparing one or more of vehicle model, scratch location, geographic location tag, or time tag for consistency with an insurance record file.

9. The computer implemented method of claim 7, wherein the method further comprises detecting a car body from an input image using object detection or segmentation methods.

10. The computer implemented method of claim 1, wherein the method further comprises retrieving a predetermined number of images based on the similarity calculation.

11. A system comprising:

at least one memory; and

at least one processor of a computer system communicatively coupled to the at least one memory, the at least one processor, responsive to instructions stored in memory, being configured to perform a method comprising;

identifying a scratch and a scratch location on a vehicle in a 3-dimensional rendering in comparison to a base image;

comparing one or more color features of the scratch in the scratch location in comparison to the scratch location in the base image;

comparing one or more texture features of the scratch in the scratch location in comparison to the scratch location in the base image; and

generating a similarity calculation based on one or more comparisons for the scratch, the scratch location, the one or more color features of the scratch, and the one or more texture features of the scratch.

12. The system of claim 11, further comprising instructions stored in memory which when executed by the at least one processor causes the at least one processor to perform the operation of confirming an insurance claim based on the similarity calculation having a score above a predetermined threshold.

13. The system of claim 11, wherein the instructions stored in memory which when executed by the at least one processor causes the at least one processor to perform the operation of identifying by providing an automobile pose estimation, a localization of the scratch location to a scratch area and a mapping to an automobile surface location by unfolding a 3-dimensional vehicle model.

14. The system of claim 11, wherein the instructions stored in memory which when executed by the at least one processor causes the at least one processor to perform the operation of comparing of one or more textures by the extraction of texture features using at least one of deep learning techniques, deep neural networks, or extreme learning machines and perform the operation of comparing of one or more color features by use of a hue, saturation, and value histogram.

15. The system of claim 11, further comprising comparing a plurality of images for consistency for vehicle model, scratch location, geographic location tag, and time tag.

16. The system of claim 15, further comprising comparing one or more of vehicle model, scratch location, geographic location tag, or time tag for consistency with an insurance record file.

17. A non-transitory computer-readable storage medium having stored therein instructions which, when executed by at least one or more processors of at least one computing device, cause a computer system to perform a method comprising:

identifying, by the at least one or more processors, a scratch and a scratch location on a vehicle in a 3-dimensional rendering in comparison to a base image;

comparing, by the at least one or more processors, one or more color features of the scratch in the scratch location in comparison to the scratch location in the base image;

comparing, by the at least one or more processors, one or more texture features of the scratch in the scratch location in comparison to the scratch location in the base image;

generating, by the at least one or more processors, a similarity calculation based on one or more comparisons for the scratch, the scratch location, the one or more color features of the scratch, or the one or more texture features of the scratch; and

presenting a result in response to the similarity calculation.

18. The non-transitory computer-readable storage medium of claim 17, further comprising confirming, by the at least one or more processors, of an insurance claim based on the similarity calculation having a score above a predetermined threshold.

19. The non-transitory computer-readable storage medium of claim 17, wherein the identifying comprises an automobile pose estimation, localization of the scratch location to a scratch area and a mapping to an automobile surface location by unfolding a 3-dimensional vehicle model.

20. The non-transitory computer-readable storage medium of claim 17, wherein the comparing of one or more texture features comprises an extraction of texture features using at least one of deep learning techniques, deep neural networks, or extreme learning machines and wherein the comparing of one or more color features comprises use of a hue, saturation, and value histogram.