Automatic identification of cloned vehicle identifiers

• US 10,255,514 B2
• Filed: 08/21/2017
• Issued: 04/09/2019
• Est. Priority Date: 08/21/2017
• Status: Active Grant

• Current Assignee: SAP SE
• Original Assignee: SAP SE
• Inventors: Yi, Chang, Luo, Qiong, Zhu, Guangji, Yu, Chao, Zhang, Qiwei, Ni, Yinling
• Primary Examiner(s): Nakhjavan, Shervin K
• Application Number:
  US15/681,931
• Publication Number:
  US 20190057275A1
• Time in Patent Office: 596 Days
• Field of Search: None
• CPC Class Codes: G06V 20/625 License plates
• CPC Class Codes: G06V 30/153 using recognition of characters or words

Abstract

Disclosed herein are system, method, and computer program product embodiments for automatically detecting cloned vehicle identifiers. An embodiment operates by determining legitimate trajectory information and illegitimate trajectory information corresponding to vehicular travel in a geographic area, and training a decision model to detect a cloned vehicle identifier based on the legitimate trajectory information and the illegitimate trajectory information. Furthermore, the embodiment identifies vehicles potentially using a cloned vehicle identifier based on the decision model and common driving patterns within the geographic area. In some embodiments, enforcement agents are notified of the vehicles potentially using a cloned vehicle identifier.

Claims

1. A computer-implemented method, comprising:

determining legitimate trajectory information associated with a plurality of legitimate vehicles;

generating illegitimate trajectory information based on the legitimate trajectory information;

training a decision model to detect a cloned vehicle identifier based on the legitimate trajectory information and the illegitimate trajectory information;

identifying sample vehicle identifier data collected by a plurality of vehicle identifier recognition devices;

determining first suspicious vehicle identifiers based at least in part on the decision model and the sample vehicle identifier data;

identifying common driving sequences between the plurality of vehicle identifier recognition devices based on the sample vehicle identifier data;

determining second suspicious vehicle identifiers based at least in part on the first suspicious vehicle identifiers and the common driving sequences; and

sending the second suspicious vehicle identifiers to an enforcement agency device associated with an enforcement agent.

2. The computer-implemented method of claim 1, wherein the determining legitimate trajectory information associated with the plurality of legitimate vehicles comprises:

for an individual vehicle of the plurality of legitimate vehicles;

receiving location data determined by a location device of the individual vehicle, the location data including a first location identifier and a first timestamp associated with a presence of the individual vehicle at a first location corresponding to the first location identifier;

receiving training vehicle identifier data determined by a vehicle identifier recognition device of the plurality of vehicle identifier recognition devices, the training vehicle identifier data including a second timestamp associated with the presence of the individual vehicle at a second location corresponding to the vehicle identifier recognition device; and

matching the location data with the training vehicle identifier data based on at least one of a difference between the first timestamp and the second timestamp being less than a predetermined temporal threshold or a distance between the first location and second location being less than a predetermined spatial threshold.

3. The computer-implemented method of claim 2, wherein the location device is a global position system device.

4. The computer-implemented method of claim 1, wherein the generating illegitimate trajectory information based on the legitimate trajectory information, comprises:

identifying a first set of vehicle identifier recognition entries of the legitimate trajectory information, the first set of vehicle identifier recognition entries corresponding to a first vehicle having a first vehicle identifier;

identifying a second set of vehicle identifier recognition entries of the legitimate trajectory information, the second set of vehicle identifier recognition entries corresponding to a second vehicle having a second vehicle identifier;

combining the first set of vehicle identifier recognition entries and the second set of vehicle identifier recognition entries to create a simulated set of vehicle identifier recognition entries; and

setting a simulated set vehicle identifier within the simulated set of vehicle identifier recognition entries to one of the first vehicle identifier or the second vehicle identifier.

5. The computer-implemented method of claim 4, wherein the generating illegitimate trajectory information based on the legitimate trajectory information, further comprises:

determining a probabilistic representation of character misrecognition by the plurality of vehicle identifier recognition devices; and

applying the probabilistic representation of character misrecognition to the simulated set of vehicle identifier recognition entries to generate the illegitimate trajectory information.

6. The computer-implemented method of claim 5, wherein the probabilistic representation of character misrecognition is an error pattern probability matrix.

7. The computer-implemented method of claim 1, wherein the decision model includes a decision tree.

8. The computer-implemented method of claim 7, wherein features of the decision tree include at least one of:

a first amount of instances a vehicle travels over predetermined speed between the plurality of vehicle identifier recognition devices;

a second amount of instances a captured vehicle identifier associated with the vehicle is recognized by the plurality of vehicle identifier recognition devices; and

a percentage of instances the captured vehicle identifier associated with the vehicle is recognized by the plurality of vehicle identifier recognition devices matches location information associated with the vehicle.

9. A non-transitory computer-readable device having instructions stored thereon that, when executed by at least one computing device, causes the at least one computing device to perform operations comprising:

determining legitimate trajectory information associated with a plurality of legitimate vehicles;

generating illegitimate trajectory information based on the legitimate trajectory information;

training a decision model to detect a cloned vehicle identifier based on the legitimate trajectory information and the illegitimate trajectory information;

identifying sample vehicle identifier data collected by a plurality of vehicle identifier recognition devices;

determining first suspicious vehicle identifiers based at least in part on the decision model and the sample vehicle identifier data;

identifying common driving sequences between the plurality of vehicle identifier recognition devices based on the sample vehicle identifier data;

determining second suspicious vehicle identifiers based at least in part on the first suspicious vehicle identifiers and the common driving sequences; and

sending the second suspicious vehicle identifiers to an enforcement agency device associated with an enforcement agent.

10. The non-transitory computer-readable device of claim 9, wherein the determining legitimate trajectory information associated with the plurality of legitimate vehicles comprises further comprises:

for an individual vehicle of the plurality of legitimate vehicles;

receiving location data determined by a location device of the individual vehicle, the location data including a first location identifier and a first timestamp associated with a first presence of the individual vehicle at a first location corresponding to the first location identifier;

receiving training vehicle identifier data determined by a vehicle identifier recognition device of the plurality of vehicle identifier recognition devices, the training vehicle identifier data including a second timestamp associated with a second presence of the individual vehicle at a second location corresponding to the vehicle identifier recognition device; and

matching the location data with the training vehicle identifier data based on at least one of a difference between the first timestamp and the second timestamp being less than a predetermined temporal threshold or a distance between the first location and second location being less than a predetermined spatial threshold.

11. The non-transitory computer-readable device of claim 10, wherein the location device is a global position system device.

12. The non-transitory computer-readable device of claim 9, wherein the generating illegitimate trajectory information based on the legitimate trajectory information, comprises:

identifying a first set of vehicle identifier recognition entries of the legitimate trajectory information, the first set of vehicle identifier recognition entries corresponding to a first vehicle having a first vehicle identifier;

identifying a second set of vehicle identifier recognition entries of the legitimate trajectory information, the second set of vehicle identifier recognition entries corresponding to a second vehicle having a second vehicle identifier;

combining the first set of vehicle identifier recognition entries and the second set of vehicle identifier recognition entries to create a simulated set of vehicle identifier recognition entries; and

setting a simulated set vehicle identifier within the simulated set of vehicle identifier recognition entries to one of the first vehicle identifier or the second vehicle identifier.

13. The non-transitory computer-readable device of claim 12, wherein the generating illegitimate trajectory information based on the legitimate trajectory information, further comprises:

determining a probabilistic representation of character misrecognition by the plurality of vehicle identifier recognition devices; and

applying the probabilistic representation of character misrecognition to the simulated set of vehicle identifier recognition entries to generate the illegitimate trajectory information.

14. The non-transitory computer-readable device of claim 13, wherein the probabilistic representation of character misrecognition is an error pattern probability matrix.

15. The non-transitory computer-readable device of claim 9, wherein the decision model includes a decision tree, and features of the decision tree include at least one of:

a first amount of instances a vehicle travels over predetermined speed between the plurality of vehicle identifier recognition devices;

a second amount of instances a captured vehicle identifier associated with the vehicle is recognized by the plurality of vehicle identifier recognition devices; and

a percentage of instances the captured vehicle identifier associated with the vehicle is recognized by the plurality of vehicle identifier recognition devices matches location information associated with the vehicle.

16. A cloned vehicle identifier detection system, comprising:

a memory; and

at least one processor coupled to the memory and configured to;

determine legitimate trajectory information associated with a plurality of legitimate vehicles;

generate illegitimate trajectory information based on the legitimate trajectory information;

train a decision model to detect a cloned vehicle identifier based on the legitimate trajectory information and the illegitimate trajectory information;

identify sample vehicle identifier data collected by a plurality of vehicle identifier recognition devices;

determine first suspicious vehicle identifiers based at least in part on the decision model and the sample vehicle identifier data;

identify common driving sequences between the plurality of vehicle identifier recognition devices based on the sample vehicle identifier data; and

determine second suspicious vehicle identifiers based at least in part on the first suspicious vehicle identifiers and the common driving sequences.

17. The cloned vehicle identifier detection system of claim 16, wherein to generate the illegitimate trajectory information based on the legitimate trajectory information, the at least one processor is further configured to:

identify a first set of vehicle identifier recognition entries of the legitimate trajectory information, the first set of vehicle identifier recognition entries corresponding to a first vehicle having a first vehicle identifier;

identify a second set of vehicle identifier recognition entries of the legitimate trajectory information, the second set of vehicle identifier recognition entries corresponding to a second vehicle having a second vehicle identifier;

combine the first set of vehicle identifier recognition entries and the second set of vehicle identifier recognition entries to create a simulated set of vehicle identifier recognition entries; and

set a simulated set vehicle identifier within the simulated set of vehicle identifier recognition entries to one of the first vehicle identifier or the second vehicle identifier.

18. The cloned vehicle identifier detection system of claim 17, wherein to generate the illegitimate trajectory information based on the legitimate trajectory information, the at least one processor is further configured to:

determine a probabilistic representation of character misrecognition by the plurality of vehicle identifier recognition devices; and

apply the probabilistic representation of character misrecognition to the simulated set of vehicle identifier recognition entries to generate the illegitimate trajectory information.

19. The cloned vehicle identifier detection system of claim 16, wherein the decision model includes a decision tree, and features of the decision tree include at least one of:

a first amount of instances a vehicle travels over predetermined speed between the plurality of vehicle identifier recognition devices;

a second amount of instances a captured vehicle identifier associated with the vehicle is recognized by the plurality of vehicle identifier recognition devices; and

a percentage of instances the captured vehicle identifier associated with the vehicle is recognized by the plurality of vehicle identifier recognition devices matches location information associated with the vehicle.

20. The cloned vehicle identifier detection system of claim 16, wherein the plurality of vehicle identifier recognition devices include a license plate recognition device.