Characterizing motion patterns of one or more agents from spatiotemporal data

Patent metadata

Field	Value
Publication number	US-9740984-B2
Application number	US-201213591079-A
Country	US
Kind code	B2
Filing date	Aug 21, 2012
Priority date	Aug 21, 2012
Publication date	Aug 22, 2017
Grant date	Aug 22, 2017

Abstract

Techniques are described to characterize motion patterns of a group of agents engaging in an activity. An analysis system receives input data associated with spatial and temporal information of at least one element of interest associated with the activity, where the object of interest may be a ball, person, animal or any other object in motion. The analysis system partitions the input data into a plurality of spatiotemporal segments and generates one or more representations of one or more sets of segments of the plurality of spatiotemporal segments based on one or more criteria. The analysis system computes a metric, such as an entropy value, for each of the one or more representations. Partial tracing data, such as ball movements in a sporting event, may be created using an inexpensive input device, such as a tablet computer, making the disclosed techniques available for a wide range of events and activities.

First claim

What is claimed is:

1. A computer-implemented method to characterize motion patterns of a first group of players engaged in a sporting activity, without requiring spatial information pertaining to any player, the computer-implemented method comprising:
receiving spatial and temporal information pertaining to a physical object in a physical playing field of the sporting activity, the sporting activity having a scoring scheme based on manipulation of the physical object in the physical playing field;
partitioning the spatial and temporal information into a plurality of spatiotemporal segments by operation of one or more computer processors, wherein each spatiotemporal segment comprises spatial and temporal information representing associated positions of the physical object in the physical playing field;
generating one or more representations of one or more sets of segments of the plurality of spatiotemporal segments, based on one or more specified criteria; and
computing, based on each of the one or more representations, a respective performance metric of the first group of players in the sporting activity, whereafter the respective performance metric is output and characterizes motion patterns of the first group of players based on spatial information pertaining to the physical object in the physical playing field and without requiring spatial information pertaining to any player.

2. The computer-implemented method of claim 1 , wherein the one or more representations comprise a probability distribution indicating a quantity of times the physical object passes through a specified region of the playing field for a given position associated with the physical object.

3. The computer-implemented method of claim 1 , wherein the performance metric measures an entropy value associated with the one or more representations.

4. The computer-implemented method of claim 1 , wherein partitioning comprises:
separating the spatial and temporal information into a group of possession strings, wherein each possession string is associated with a time interval when the first group of players was in a winning position relative to a second group of players and according to the scoring scheme.

5. The computer-implemented method of claim 1 , wherein partitioning comprises:
separating the spatial and temporal information into a group of possession strings, wherein each possession string comprises a different set of time steps from a plurality of time steps during the sporting activity, wherein each possession string is associated with a time interval when the first group of players had possession of the physical object;
populating each time step with a position of the physical object at the associated time step; and
extracting one or more spatiotemporal segments from each possession string, wherein the one or more spatiotemporal segments are included in the plurality of spatiotemporal segments.

6. The computer-implemented method of claim 5 , wherein each time step in the set of time steps is of equal duration.

7. The computer-implemented method of claim 5 , wherein the time steps in each segment of the one or more segments are sequential.

8. The computer-implemented method of claim 1 , further comprising performing an analysis based on the performance metric.

9. The computer-implemented method of claim 8 , further comprising receiving a query associated with the sporting activity, wherein the analysis is performed responsive to the query.

10. The computer-implemented method of claim 1 , further comprising modifying the performance metric based on one or more statistics associated with the sporting activity.

11. The computer-implemented method of claim 1 , wherein the spatial and temporal information further pertains to a plurality of time steps during the sporting activity, wherein the spatial and temporal information is partitioned into the plurality of spatiotemporal segments using a sliding window;
wherein each spatiotemporal segment in the plurality of spatiotemporal segments has a predefined first time duration of the sliding window, wherein the time period of each spatiotemporal segment overlaps the time periods of one or more other spatiotemporal segments;
wherein each spatiotemporal segment comprises spatial and temporal information representing associated positions of the physical object in the physical playing field, from an equal number of the time steps which when grouped together sum to the predefined first time duration.

12. The computer-implemented method of claim 11 , wherein the sporting activity comprises a team-based, continuous, non-discretizable sporting activity, wherein the first group comprises a first of two teams involved in the sporting activity, wherein each group comprises a respective team, wherein the physical object is selected from a ball, a puck, and a shuttle;
wherein the physical object is a ball, wherein one or more performance measurements, which include the respective performance measurement of each of the one or more representations, specifies: (i) whether the team's performance during a particular match is consistent with performance during prior matches; (ii) whether the team's play varies depending on which team is on the opposing side; (iii) whether the behavior is influenced by coaching staff; and (iv) whether the behavior is influenced by one or more key players.

13. The computer-implemented method of claim 12 , wherein the one or more representations comprise a probability distribution indicating a quantity of times the physical object passes through a specified region of the playing field for a given position associated with the physical object, wherein the performance metric measures an entropy value associated with the one or more representations.

14. The computer-implemented method of claim 13 , wherein partitioning comprises, in a first instance:
separating the spatial and temporal information into a group of possession strings, wherein each possession string is associated with a time interval when the first group of players was in a winning position relative to a second group of players and according to the scoring scheme.

15. The computer-implemented method of claim 14 , wherein partitioning comprises, in a second instance:
separating the spatial and temporal information into a group of possession strings, wherein each possession string comprises a different set of time steps from the plurality of time steps, wherein each possession string is associated with a time interval when the first group of players had possession of the physical object;
populating each time step with a position of the physical object at the associated time step; and
extracting one or more spatiotemporal segments from each possession string, wherein the one or more spatiotemporal segments are included in the plurality of spatiotemporal segments.

16. The computer-implemented method of claim 15 , wherein the sporting activity comprises soccer, wherein the ball is a soccer ball, wherein each time step in the set of time steps is of equal duration, wherein the time steps in each segment of the one or more segments are sequential, wherein the computer-implemented method further comprises:
modifying the performance metric based on one or more statistics associated with the sporting activity;
receiving an input query associated with the sporting activity;
responsive to the input query, performing an analysis based on the performance metric; and
generating a visual representation based on the analysis, whereafter the visual representation is output.

17. A non-transitor

Assignees

• Lucey Patrick
• Bialkowski Alina
• Matthews Iain
• Carr G Peter
• Foote Eric
• Disney Entpr Inc

Inventors

• Lucey Patrick
• Bialkowski Alina
• Matthews Iain
• Carr G Peter
• Foote Eric

Classifications

• G06N5/043 Primary

  Distributed expert systems; Blackboards · CPC title