Method and system for detecting uninsured motor vehicles

• US 9,704,201 B2
• Filed: 07/30/2014
• Issued: 07/11/2017
• Est. Priority Date: 07/30/2014
• Status: Active Grant

• Current Assignee: Conduent Business Services, LLC (Conduent, Inc.)
• Original Assignee: Conduent Business Services, LLC (Conduent, Inc.)
• Inventors: Burry, Aaron M., Bulan, Orhan, Jackson, David Martin Todd, Islam, Abu S., Paul, Peter
• Primary Examiner(s): Park, Soo Jin
• Application Number:
  US14/446,938
• Publication Number:
  US 20160035037A1
• Time in Patent Office: 1,077 Days
• Field of Search: None
• CPC Class Codes: G06Q 40/08 Insurance
• CPC Class Codes: G06V 10/60 relating to illumination properties, e.g. using a reflectance or lighting model
• CPC Class Codes: G06V 20/63 Scene text, e.g. street names

Abstract

A video sequence can be continuously acquired at a predetermined frame rate and resolution by an image capturing unit installed at a location. A video frame can be extracted from the video sequence when a vehicle is detected at an optimal position for license plate recognition by detecting a blob corresponding to the vehicle and a virtual line on an image plane. The video frame can be pruned to eliminate a false positive and multiple frames with respect to a similar vehicle before transmitting the frame via a network. A license plate detection/localization can be performed on the extracted video frame to identify a sub-region with respect to the video frame that are most likely to contain a license plate. A license plate recognition operation can be performed and an overall confidence assigned to the license plate recognition result.

Claims

1. A method for detecting an uninsured motor vehicle, said method comprising:

continuously acquiring a video sequence at a predetermined frame rate and resolution by an image-capturing unit installed at a location to extract a video frame from said video sequence when a vehicle is detected at an optimal position for license plate recognition, wherein said image-capturing unit communicates with a video processing unit through a computer network, said video processing unit comprising a video frame triggering unit;

detecting and localizing said license plate on said extracted video frame to identify a sub-region in a region of interest of said video frame that includes a license plate;

performing a license plate recognition with respect to said license plate detected and localized on said extracted video frame utilizing a character level segmentation and an optical character recognition and assigning an overall confidence to a license plate recognition result with respect to said extracted video frame, said overall confidence indicative of a likelihood that said extracted video frame contains said license plate and that said extracted video frame is capable of eliminating at least some false alarms arising from said video frame triggering unit, and wherein said overall confidence is assigned based on a success of said character level segmentation and said optical character recognition; and

identifying insurance with respect to a detected vehicle from a vehicle insurance database after said detecting and localizing said license plate on said extracted video frame and after said performing said license plate recognition with respect to said license, plate detected and localized on said extracted video frame, and thereafter automatically sending a notification/ticket to a registrant of said vehicle, if said vehicle is identified as uninsured.

2. The method of claim 1 further comprising pruning said video frame to eliminate a false positive and multiple frames with respect to a vehicle similar to said vehicle detected at said optimal position before transmitting said video frame via a network and prior to said detecting and localizing said license plate on said extracted video frame.

3. The method of claim 2 further comprising extracting said video frame by transferring a frame of interest to a central processing unit for further processing and wherein said video frame triggering unit accomplishes video triggering utilizing a plurality of virtual lines, said vide processing unit comprising said central processing unit.

4. The method of claim 3 further comprising determining said frame rate and resolution based on requirements of said video frame triggering unit and a license plate recognition unit and wherein said plurality of virtual lines comprise two virtual lines defined on image plane.

5. The method of claim 4 wherein said image capturing unit comprises an RGB image capturing unit and a near infra-red image capturing unit and wherein said computer network includes a collection of networks and gateways that utilize a Transmission Control Protocol/Internet Protocol (TCP/IP) suite of protocols to communicate with one another.

6. The method of claim 4 further comprising obtaining said database from a department of motor vehicle and from an insurance company.

7. The method of claim 3 wherein said detecting and localizing said license plate on said extracted video frame, further comprises:

computing a virtual trap to detect said vehicle at a specific position in said image capturing unit view; and

defining said trap by a virtual line, a polygon, and said plurality of virtual lines and a plurality of virtual polygons.

8. The method of claim 7 further comprising performing said motion detection by a temporal difference approach, a pixel-level optical flow approach, and a block-matching algorithm.

9. The method of claim 3 further comprising eliminating said false positive due to a cast shadow utilizing a machine learning approach and a shadow suppression technique.

10. The method of claim 2 further comprising extracting said video frame in said central processing unit after transferring said video sequence via said computer network, wherein said computer network is operably connected to said video processing unit that comprises a video acquisition unit, said video frame triggering unit, and a video frame pruning unit.

11. The method of claim 3 wherein said detecting and localizing said license plate on said extracted video frame, further comprises:

detecting a blob corresponding to said vehicle on an image plane of said region of interest utilizing a background subtraction and a motion detection technique in which motion regions in said video frame are extracted utilizing at least one of a pixel-level optical flow method or a block-matching algorithm; and

computing motion detection with said motion detection technique utilizing a temporal different method in conjunction with a morphological operation for vehicle detection in said region of interest, wherein said temporal difference method comprises double frame difference.

12. The method of claim 11 wherein said background subtraction, further comprises:

highlighting an object in said video sequence when a static image-capturing unit is employed to capture said video sequence;

computing an absolute intensity/color difference between a known background image and each image in said video sequence by a background removal when an image of a background without any foreground object is available; and

classifying a pixel for which a computed distance in an intensity/color space is small as a background pixel.

13. The method of claim 12 further comprising performing said motion detection by a temporal difference approach, a pixel-level optical flow approach, and a block-matching algorithm.

14. The method of claim 11 further comprising:

applying a shadow removal technique to remove a shadow portion of said blob detected utilizing said background subtraction and said motion detection technique, said blob comprising a vehicle blob and wherein said blob is processed by said video frame triggering unit; and

determining whether said license plate recognition result is a candidate for an automated processing and requires a manual validation by said confidence score.

15. The method of claim 2 further comprising attaching said video frame extracted to said ticket as an evidence to prove that said vehicle is driving in traffic when uninsured.

16. A system for detecting an uninsured motor vehicle, said system comprising:

at least one processor; and

memory comprising instructions stored therein, which when executed by said at least one processor, causes said at least one processor to perform operations comprising;

continuously acquiring a video sequence at a predetermined frame rate and resolution by an image-capturing unit installed at a location to extract a video frame from said video sequence when a vehicle is detected at an optimal position for license plate recognition, wherein said image-capturing unit communicates with a video processing unit through a computer network, said video processing unit comprising a video frame triggering unit;

detecting and localizing said license plate on said extracted video frame to identify a sub-region in a region of interest of said video frame that includes a license plate;

performing a license plate recognition with respect to said license plate detected and localized on said extracted video frame utilizing a character level segmentation and an optical character recognition and assigning an overall confidence to a license plate recognition result with respect to said extracted video frame, said overall confidence indicative of a likelihood that said extracted video frame contains said license plate and that said extracted video frame is capable of eliminating at least some false alarms arising from said video frame triggering unit, and wherein said overall confidence is assigned based on a success of said character level segmentation and said optical character recognition; and

identifying insurance with respect to a detected vehicle from a vehicle insurance database after said detecting and localizing said license plate on said extracted video frame and after said performing said license plate recognition with respect to said license plate detected and localized on said extracted video frame, and thereafter automatically sending a notification/ticket to a registrant of said vehicle, if said vehicle is identified as uninsured.

17. The system of claim 16 wherein said operations further comprise pruning said video frame to eliminate a false positive and multiple frames with respect to a vehicle similar to said vehicle detected at said optimal position before transmitting said video frame via a network and prior to said detecting and localizing said license plate on said extracted video frame.

18. The system of claim 17 wherein said operations further comprise:

extracting said video frame by transferring a frame of interest to a central processing unit for further processing and wherein said video frame triggering unit accomplishes video triggering utilizing a plurality of virtual lines; and

wherein said operations for detecting and localizing said license plate on said extracted video frame further comprises operations for;

detecting a blob corresponding to said vehicle on an image plane of said region of interest utilizing a background subtraction and a motion detection technique in which motion regions in said video frame are extracted utilizing at least one of a pixel-level optical flow method or a block-matching algorithm; and

computing motion detection with said motion detection technique utilizing a temporal different method in conjunction with a morphological operation for vehicle detection in said region of interest, wherein said temporal difference method comprises double frame difference.

19. The system of claim 17 wherein said operations further comprise:

extracting said video frame in said central processing unit after transferring said video sequence via said computer network, wherein said computer network is configured to be operably connected to said video processing unit that comprises a video acquisition unit, said video frame triggering unit, and a video frame pruning unit.

20. A non-transitory machine-readable medium comprising instructions stored therein, which when executed by a machine, cause the machine to perform operations comprising:

continuously acquiring a video sequence at a predetermined frame rate and resolution by an image-capturing unit installed at a location to extract a video frame from said video sequence when a vehicle is detected at an optimal position for license plate recognition, wherein said image-capturing unit communicates with a video processing unit through a computer network, said video processing unit comprising a video frame triggering unit;

detecting and localizing said license plate on said extracted video frame to identify a sub-region in a region of interest of said video frame that includes a license plate;

performing a license plate recognition with respect to said license plate detected and localized on said extracted video frame utilizing a character level segmentation and an optical character recognition and assigning an overall confidence to a license plate recognition result with respect to said extracted video frame, said overall confidence indicative of a likelihood that said extracted video frame contains said license plate and that said extracted video frame is capable of eliminating at least some false alarms arising from a video frame triggering unit, and wherein said overall confidence is assigned based on a success of said character level segmentation and said optical character recognition; and

identifying insurance with respect to a detected vehicle from a vehicle insurance database after detecting and localizing said license plate on said extracted video frame and after performing said license plate recognition with respect to said license plate detected and localized on said extracted video frame, and thereafter automatically sending a notification/ticket to a registrant of said vehicle, if said vehicle is identified as uninsured.