VIDEO OBJECT DATA STORAGE AND PROCESSING SYSTEM

• US 20170300754A1
• Filed: 04/13/2017
• Published: 10/19/2017
• Est. Priority Date: 04/14/2016
• Status: Active Grant

• Current Assignee: KickView Corporation
• Original Assignee: KickView Corporation
• Inventors: Ohm, David R., North, Robert L., Muchmore, Kyle A.
• Granted Patent: US 10,217,001 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 18/2414 Smoothing the distance, e.g. radial basis function networks [RBFN]
• CPC Class Codes: G06F 18/24317 Piecewise classification, i.e. whereby each classification requires several discriminant rules
• CPC Class Codes: G06F 18/285 Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system
• CPC Class Codes: G06V 10/764 using classification, e.g. of video objects
• CPC Class Codes: G06V 10/82 using neural networks
• CPC Class Codes: G06V 10/87 using selection of the recognition techniques, e.g. of a classifier in a multiple classifier system
• CPC Class Codes: G06V 20/41 Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items segmenting video sequences G06V20/49
• CPC Class Codes: G06V 20/70 Labelling scene content, e.g. deriving syntactic or semantic representations

Abstract

A video object data storage and display system comprising a video object data selection and viewing portion and a video object data storage portion. The system comprises a video object having a: scale/size, pose/tilt, location, and frame/time. The system further comprises a database.

Claims

1. A video processing system comprising,

an end-user device displaying one or more sections of video, the one or more sections of video comprising one or more video objects;

one or more video object storing devices, wherein information related to the one or more video objects is located on the one or more video object storing devices;

one or more object classification devices, the one or more classification devices implementing a neural network to recognize and identify the one or more video objects;

a communication control device processing communications between the end-user device, one or more video object storing devices, and one or more classification devices.

2. The video processing system of claim 1 wherein, the one or more video objects comprises at least one of,

a person,

an animal,

an activity,

a motion,

an object, and

an event.

3. The video processing system of claim 1 wherein, implementing a neural network to recognize and identify the one or more video objects comprises utilizing an algorithm associated with a network structure to associate one or more AI tags with each of the one or more video objects.

4. The video processing system of claim 3 wherein,

the one or more object classification devices comprises a first object classification device and a second object classification device;

the first object classification device comprises a first algorithm and a first network structure; and

the second object classification device comprises a second algorithm and a second network structure, wherein, at least one of the second algorithm and the second network structure is different than the first algorithm and first network structure.

5. The video processing system of claim 4, wherein,

the first classification device associates one or more first AI tags with each of the one or more video objects;

the second classification device associates one or more second AI tags with each of the one or more video objects; and

at least one of the second AI tags is different than the first AI tags.

6. The video processing system of claim 5 wherein, the second classification device verifies the accuracy of the AI tags from the first verification device.

7. The video processing system of claim 6 wherein, the communication control device verifies the accuracy of the AI tags from the first verification device by comparing the AI tags from the second verification device with user-supplied metadata.

8. A method of identifying a video object comprising,

selecting at least one video object in at least one frame of an identified video;

using at least one of a first classification device and a second classification device to recognize the at least one video object, wherein,

the first classification device comprises a first neural network having,

a first network structure, and

a first algorithm,

the second classification device comprises a second neural network having,

a second network structure, and

a second algorithm;

associating the at least one recognized video object with at least one word, wherein, the at least one word is related to at least one feature of the at least one recognized video object.

9. The method of claim 8 wherein, the second neural network is different than the first neural network.

10. The method of claim 8 wherein,

the identified video comprises a plurality of videos;

and further comprising,

using the at least one word to locate one or more additional video objects in the plurality of videos; and

further comprising,

associating metadata with the at least one video object after selecting the at least one video object.

11. The method of claim 10 wherein,

selecting at least one video object in at least one frame of an identified video comprises using a pointing device to selectively identify an object;

and

associating metadata with the at least one video object comprises at least one of manually,

giving the object a name,

selecting a category for the object, and

typing keywords related to the object.

12. The method of claim 11, wherein using a pointing device to selectively identify an object comprises using a computing mouse to create a box around the object.

13. A video object data storage and display system comprising,

a video object data selection portion comprising,

a video object scale,

a video object pose,

a video object location, and

a video object frame;

and

a video object data storage portion comprising a database.

14. The video object data storage and display system of claim 13 wherein,

the video object scale, video object pose, and video object location:

comprise portions of a video object shape for a plurality of predetermined video frames, and

determine the shape, position, and orientation of a video object; and

the video object frame comprises a time.

15. The video object data storage and display system of claim 14 wherein,

the video object shape comprises,

a major axis having a major axis length,

a minor axis having a minor axis length;

the video object position comprises a center-mass location;

the video object orientation comprises an angle; and

the database comprises a first vector, the first vector comprises a plurality of first vector parameters, the plurality of first vector parameters comprising,

a first major axis length,

a first minor axis length,

a first center-mass location,

a first angle, and

a first time.

16. The video object data storage and display system of claim 15, wherein,

the database comprises a second vector, the second vector comprises a plurality of second vector parameters, the plurality of second vector parameters comprising,

a second major axis length,

a second minor axis length,

a second center-mass location,

a second angle, and

a second time;

and

the database comprises a third vector, the third vector comprises a plurality of third vector parameters, the plurality of third vector parameters comprising,

a third major axis length,

a third minor axis length,

a third center-mass location,

a third angle; and

the third vector parameters are obtained by interpolating the first vector parameters and the second vector parameters for a given third time, wherein the third time is greater than the first time and less than the second time.

17. The video object data storage and display system of claim 16, wherein, one of a Linear Least Squares and a Linear Weighted Least Squares approach is used to obtain the third vector parameters.

18. The video object data storage and display system of claim 15, wherein,

the center-mass location comprises a location of an intersection of the major axis and the minor axis;

and

the angle comprises a distance between the major axis and a reference axis.

19. The video object data storage and display system of claim 18 wherein,

the reference axis comprises the x-axis;

the frame comprises a value associated with an object clarity and an object pose; and

further comprising,

at least one display device comprising,

a video display portion, and

a video object interaction portion.

20. A non-transitory, tangible processor readable storage medium, encoded with processor executable code to perform a method for storing the location of an object through the scene of a video comprising,

obtaining an object location for a plurality of video frames;

enclosing the object in the plurality of video frames with a geometric shape;

obtaining parameters related to the ellipse for the plurality of video frames; and

using at least some the parameters for at least two of the plurality of video frames to obtain the object location for another video frame.