Orientation estimation method and orientation estimation device

Patent metadata

Field	Value
Publication number	US-10121260-B2
Application number	US-201515329031-A
Country	US
Kind code	B2
Filing date	Oct 23, 2015
Priority date	Oct 30, 2014
Publication date	Nov 6, 2018
Grant date	Nov 6, 2018

Abstract

A processor of an orientation estimation device, as an orientation estimation method for estimating an orientation of a player existing in a sport video, receives reference orientation information which is information obtained based on information input by a user and which is information designating a joint position of a specific player existing in the sport video of an estimation target game and estimates an orientation of an estimation target player who is a player other than the specific player existing in a sport video of an estimation target using the reference orientation information.

First claim

The invention claimed is:

1. An orientation estimation method for estimating an orientation of a player existing in a sport video by a processor, wherein the processor
receives reference orientation information of a specific player included in the sport video, the reference orientation information designating a plurality of joint positions of the specific player;
applies the reference orientation information to an estimation target player that is included in the sport video and different from the specific player; and estimates an orientation of the estimation target player using the reference orientation,
wherein information designating the joint positions of the specific player is generated based on user's input individually performed for at least a plurality of places of joints, and
wherein information specifying the estimation target player is generated based on the user's input individually performed for a number of places fewer than a number of the plurality of places of joints.

2. The orientation estimation method according to claim 1 , wherein the processor further estimates information about a color of the estimation target player, and information about a color of the joint positions positions of the specific player.

3. The orientation estimation method according to claim 1 ,
wherein a sport of the sport video is a team competition sport, and
the reference orientation information is information input for a player of each of at least two opposing teams.

4. The orientation estimation method according to claim 3 ,
wherein the processor estimates the orientation of the estimation target player within a team among the at least two opposing teams using the reference orientation information for each team of the at least two opposing teams.

5. The orientation estimation method according to claim 1 ,
wherein a sport of the sport video is a certain sport in which the specific player and the estimation target player wear prescribed uniforms for each team of the at least two opposing teams.

6. The orientation estimation method according to claim 1 ,
wherein the processor specifies the estimation target player by information input from the user.

7. An orientation estimation device, comprising: a processor,
wherein the processor receives reference orientation information of a specific player included in the sport video, the reference orientation information designating a plurality of joint positions of the specific player, and
applies the reference orientation information to an estimation target player that is included in the sport video and different from the specific player; and
estimates an orientation of the estimation target player using the reference orientation information,
wherein information designating the joint positions of the specific player is generated based on user's input individually performed for at least a plurality of places of joints, and
wherein information specifying the estimation target player is generated based on the user's input individually performed for a number of places fewer than a number of the plurality of places of joints.

8. The orientation estimation device according to claim 7 ,
wherein the processor further estimates information about a color of the estimation target player, and information about a color of the joint positions of the specific player.

9. The orientation estimation device according to claim 7 ,
wherein a sport of the sport video is a team competition sport, and
the reference orientation information is information input for a player of each of at least two opposing teams.

10. The orientation estimation device according to claim 9 ,
wherein the processor estimates the orientation of the estimation target player within a team among the at least two opposing teams using the reference orientation information for each team of the at least two opposing teams.

11. The orientation estimation device according to claim 7 ,
wherein a sport of the sport video is a certain sport in which the specific player and the estimation target player wear prescribed team uniforms for each team of the at least two opposing teams.

12. The orientation estimation device according to claim 7 ,
wherein the processor specifies the estimation target player by information input through the input device from the user.

13. The orientation estimation method according to claim 1 ,
wherein the joint positions include position includes any of positions of a head, a chest, a waist, a part above knee, and a part below knee of a player.

14. The orientation estimation device according to claim 7 ,
wherein the joint positions include any of positions of a head, a chest, a waist, a part above knee, and a part below knee of a player.

Assignees

• Panasonic Ip Man Co Ltd

Inventors

• Kawaguchi Kyoko
• Sekii Taiki

Classifications

• G06V10/764 Primary

  using classification, e.g. of video objects · CPC title

• G06T7/73 Primary

  using feature-based methods · CPC title

• G06T2207/10016

  Video; Image sequence · CPC title

• G06F18/24133

  Distances to prototypes · CPC title

• G06V10/50

  by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis · CPC title