Method of extracting ridge data and apparatus and method for tracking joint motion of object

• US 9,460,364 B2
• Filed: 12/08/2014
• Issued: 10/04/2016
• Est. Priority Date: 12/09/2013
• Status: Active Grant

• Current Assignee: Postech Academy-Industry Foundation
• Original Assignee: Postech Academy-Industry Foundation
• Inventors: Kim, Yeon Ho, Kim, Dai Jin
• Primary Examiner(s): Patel, Nirav G
• Assistant Examiner(s): SHIN, BRIAN D
• Application Number:
  US14/562,848
• Publication Number:
  US 20150161445A1
• Time in Patent Office: 666 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06V 10/752 Contour matching
• CPC Class Codes: G06V 40/23 Recognition of whole body movements, e.g. for sport training

Abstract

A method of extracting ridge data includes selecting a point in the contour of an object from an image of the object, calculating a chamfer distance of the point in the contour, and determining whether the point in the contour is a ridge point based on relationships between the chamfer distance and chamfer distances of points in the contour adjacent to the point in the contour. A method of tracking a joint motion of an object includes acquiring initial information of the object including information on at least one ridge region representing a distribution of ridge points of the object, and tracking the joint motion of the object from temporally consecutive images of the object based on the initial information.

Claims

1. A method of extracting ridge data in a method of tracking motion of an object, the method comprising:

selecting a point in a contour of the object from an image of the object;

calculating a chamfer distance of the point in the contour;

determining whether the point in the contour is a ridge point based on relationships between the chamfer distance and chamfer distances of points in the contour adjacent to the point in the contour; and

tracking a position of a specific part of the object based on coordinates of a ridge point in one image among temporally consecutive images of the object and a chamfer distance of the ridge point,

wherein a distribution correction factor representing a distribution of chamber distances is set to vary according a part of the object from which ridge data is acquired.

2. The method of claim 1, wherein a point in the contour satisfying an expression below is selected according to the relationships between the chamfer distance and the chamfer distances of the points in the contour adjacent to the point in the contour:

3. A method of tracking a joint motion of an object, the method comprising:

acquiring initial information of the object including information on at least one ridge region representing a distribution of ridge points of the object; and

tracking the joint motion of the object based on coordinates of a ridge point in one mage among temporally consecutive images of the object and a chamfer distance of the ridge point, and based on the initial information,

wherein a distribution correction factor representing a distribution of chamber distances is set to vary according to a part of the object from which ridge data is acquired.

4. The method of claim 3, wherein the acquiring of the initial information comprises:

selecting a point in a contour of the object from an image of the object in a predetermined posture;

calculating a chamfer distance of the point in the contour;

selecting at least one object ridge point satisfying a predetermined ridge condition of a chamfer distance and a distribution correction factor representing a distribution of chamfer distances among points in the contour; and

determining a group consisting of the at least one object ridge point as the at least one ridge region.

5. The method of claim 4, wherein a point in the contour satisfying an expression below is selected under the predetermined ridge condition:

6. The method of claim 4, wherein the acquiring of the initial information comprises finding two linear regression equations representing an upper arm and a fore arm by performing linear regression on ridge points in an upper arm region and a fore arm region among the at least one ridge region, and determining a cross point of the two linear regression equations as an elbow position.

7. The method of claim 4, wherein the acquiring of the initial information comprises:

finding two linear regression equations representing a thigh and a lower leg by performing linear regression on ridge points in a thigh region and a lower leg region among the at least one ridge region;

determining a cross point of the two linear regression equations as a knee position when an acute angle between crossing angles formed by the two linear regression equations is larger than a predetermined angle; and

determining a middle point between coordinates of a hip joint and floor plane coordinates closest to the hip joint coordinates as the knee position when the acute angle is equal to or smaller than the predetermined angle.

8. The method of claim 3, wherein the tracking of the joint motion comprises:

a head tracking operation of determining a position of a head center in a t^th image among the temporally consecutive images based on coordinates of at least one head ridge point satisfying a predetermined head condition among the object ridge points;

a torso tracking operation of determining torso joint positions in the t^th image based on coordinates of at least one torso ridge point satisfying a predetermined torso condition among the object ridge points other than the at least one head ridge point; and

a limb tracking operation of determining limb joint positions in the t^th image based on coordinates of at least one limb ridge point satisfying a predetermined limb condition among the object ridge points other than the at least one head ridge point and the at least one torso ridge point.

9. The method of claim 8, wherein the predetermined head condition is expressed as at least one equation or inequality relating a displacement of the head center in a (t−

1)^th image among the temporally consecutive images and a head ridge region in the t^th image.

10. The method of claim 8, wherein the predetermined torso condition is expressed as at least one equation or inequality relating a displacement of a torso center in a (t−

1)^th image among the temporally consecutive images and the head center position in the t^th image.

11. The method of claim 8, wherein a position of a torso center among the torso joint positions is determined based on head-shoulder-torso (HST) structure information in the initial information, and

the HST structure information is expressed as distances and angles between the head center, both shoulders, and the torso center.

12. The method of claim 8, wherein positions of both elbows and positions of both knees among the limb joint positions are determined based on positions of both shoulders and both hip joints, which are upper joints of the elbows and the knees, respectively, and

positions of both hands and positions of both feet among the limb joint positions are determined based on positions of both the elbows and both the knees, which are upper joints of the hands and the feet, respectively.

13. The method of claim 8, wherein the predetermined limb condition is expressed as at least one equation or inequality relating a Euclidean distance and a geodesic distance between the at least one limb ridge point in a region of a lower joint and an upper joint directly above the lower joint.

14. The method of claim 13, wherein the limb tracking operation comprises, when a number of the at least one limb ridge point satisfying the predetermined limb condition is equal to or smaller than a predetermined number, calculating the limb joint positions based on the initial information, a position of the upper joint in a (t−

1)^th image, and a position of a (t−

1)^th limb joint.

15. The method of claim 14, wherein the at least one limb ridge point is collected after ridge points of other ridge regions are removed using a data slicing technique.

16. An apparatus for tracking a joint motion of an object, the apparatus comprising:

an initial information acquisition unit configured to acquire initial information of the object including information on at least one ridge region representing a distribution of ridge points of the object; and

a joint motion tracking unit configured to track the joint motion of the object based on coordinates of a ridge point in one image among temporally consecutive images of the object and a chamfer distance of the ridge point, and based on the initial information,

wherein a distribution correction factor representing a distribution of chamfer distances is set.

17. The apparatus of claim 16, wherein the initial information acquisition unit comprises:

a point selector configured to select a point in a contour of the object from an image of the object in a predetermined posture;

a slope calculator configured to calculate a chamfer distance of the point in the contour;

a ridge point determiner configured to select at least one object ridge point satisfying a predetermined ridge condition of a chamfer distance and a distribution correction factor representing a distribution of chamfer distances among points in the contour; and

a ridge region extractor configured to determine a group consisting of the at least one object ridge point as the at least one ridge region.

18. The apparatus of claim 17, wherein a point in the contour satisfying an expression below selected under the predetermined ridge condition:

19. The apparatus of claim 17, wherein the joint motion tracking unit comprises:

a head tracker configured to determine a head center position in a t^th image among the temporally consecutive images based on coordinates of at least one head ridge point satisfying a predetermined head condition among the object ridge points;

a torso tracker configured to determine torso joint positions in the t^th image based on coordinates of at least one torso ridge point satisfying a predetermined torso condition among the object ridge points other than the at least one head ridge point; and

a limb tracker configured to determine limb joint positions in the t^th image based on coordinates of at least one limb ridge point satisfying a predetermined limb condition among the object ridge points other than the at least one head ridge point and the at least one torso ridge point.