Iterative closest point technique based on a solution of inverse kinematics problem

Patent metadata

Field	Value
Publication number	US-9747717-B2
Application number	US-201514749295-A
Country	US
Kind code	B2
Filing date	Jun 24, 2015
Priority date	May 13, 2015
Publication date	Aug 29, 2017
Grant date	Aug 29, 2017

Abstract

Techniques related to non-rigid transformations for articulated bodies are discussed. Such techniques may include repeatedly selecting target positions for matching a kinematic model of an articulated body, generating virtual end-effectors for the kinematic model and corresponding to the target positions, generating an inverse kinematics problem including a Jacobian matrix, and determining a change in kinematic model parameters based on the inverse kinematics problem until a convergence is attained.

First claim

What is claimed is:

1. A method for providing a non-rigid transformation for an articulated body comprising:
selecting, based on input image data, a plurality of target positions for matching a kinematic model representing an articulated body, wherein the kinematic model comprises a pose based on initial kinematic model parameters that provide spatial relationships of elements of the kinematic model, and wherein the kinematic model comprises a plurality of joints, a plurality of end-effectors, and links between selected joints and end-effectors all within a model skin of the kinematic model;
generating, in addition to the end-effectors of the kinematic model, a plurality of virtual end-effectors corresponding to the target positions, wherein each of the virtual end-effectors is generated at a point on the model skin of the kinematic model closest to an associated target position of the target positions;
generating an inverse kinematics problem comprising a Jacobian matrix based on the initial kinematic model parameters, the target positions, and the virtual end-effectors;
determining a change in the kinematic model parameters based on the inverse kinematics problem;
repeating the selecting the plurality of target positions, generating the plurality of virtual end-effectors, generating the inverse kinematics problem, and determining the change in the kinematic model parameters until a convergence is attained, wherein repeating the generating the plurality of virtual end-effectors comprises determining new virtual end-effectors on the model skin of the kinematic model closest to associated target positions at each iteration; and
outputting resultant kinematic model parameters associated with the convergence.

2. The method of claim 1 , wherein the kinematic model parameters comprise at least one of an angle of rotation for a first joint or a translation distance for a second joint.

3. The method of claim 1 , wherein a first virtual end-effector is associated with a first joint of the kinematic model by a virtual link.

4. The method of claim 1 , wherein determining the change in the kinematic model parameters comprises determining the change in the kinematic model parameters that minimize the inverse kinematics problem.

5. The method of claim 1 , wherein the inverse kinematics problem comprises at least one first kinematic model parameter comprising a feasibility set such that the first kinematic model parameter must be within the feasibility set.

6. The method of claim 1 , wherein the Jacobian matrix comprises at least one element having a target weighting parameter associated with a first target position of the plurality of target positions.

7. The method of claim 1 , wherein the Jacobian matrix comprises at least one element having a joint weighting parameter associated with a first joint of the elements of the kinematic model.

8. The method of claim 1 , wherein the Jacobian matrix comprises at least one element having a repulsive target functionality associated with a first target position of the plurality of target positions.

9. The method of claim 1 , wherein the input image data comprises at least one of a 3D point cloud or a depth map.

10. The method of claim 1 , wherein repeating selecting the plurality of target positions comprises randomly selecting a new plurality of target positions based on the input image data at each iteration.

11. The method of claim 1 , wherein the articulated body represents at least one of a hand, a human body, an animal body, a machine, a device, a laptop, a closet, or a robot.

12. A system for providing a non-rigid transformation for an articulated body comprising:
a memory to store image data; and
a central processor coupled to the memory, the central processor to select, based on input image data, a plurality of target positions for matching a kinematic model representing an articulated body, wherein the kinematic model comprises a pose based on initial kinematic model parameters that provide spatial relationships of elements of the kinematic model, and wherein the kinematic model comprises a plurality of joints, a plurality of end-effectors, and links between selected joints and end-effectors all within a model skin of the kinematic model, to generate, in addition to the end-effectors of the kinematic model, a plurality of virtual end-effectors corresponding to the target positions, wherein each of the virtual end-effectors is generated at a point on the model skin of the kinematic model closest to an associated target position of the target positions, to generate an inverse kinematics problem comprising a Jacobian matrix based on the initial kinematic model parameters, the target positions, and the virtual end-effectors, to determine a change in the kinematic model parameters based on the inverse kinematics problem, and to repeat the selection of the plurality of target positions, generation of the plurality of virtual end-effectors, generation of the inverse kinematics problem and determination of the change in the kinematic model parameters until a convergence is attained, wherein to repeat the generation of the plurality of virtual end-effectors comprises the central processor to determine new virtual end-effectors on the model skin of the kinematic model closest to associated target positions at each iteration, and to output resultant kinematic model parameters associated with the convergence.

13. The system of claim 12 , wherein the inverse kinematics problem comprises at least one first kinematic model parameter comprising a feasibility set such that the first kinematic model parameter must be within the feasibility set.

14. The system of claim 12 , wherein the Jacobian matrix comprises at least one element having a target weighting parameter associated with a first target position of the plurality of target positions.

15. The system of claim 12 , wherein the Jacobian matrix comprises at least one element having a joint weighting parameter associated with a first joint of the elements of the kinematic model.

16. The system of claim 12 , wherein the Jacobian matrix comprises at least one element having a repulsive target functionality associated with a first target position of the plurality of target positions.

17. At least one non-transitory machine readable medium comprising a plurality of instructions that, in response to being executed on a computing device, cause the computing device to provide a non-rigid transformation for an articulated body by:
selecting, based on input image data, a plurality of target positions for matching a kinematic model representing an articulated body, wherein the kinematic model comprises a pose based on initial kinematic model parameters that provide spatial relationships of elements of the kinematic model, and wherein the kinematic model comprises a plurality of joints, a plurality of end-effectors, and links between selected joints and end-effectors all within a model skin of the kinematic model;
generating, in addition to the end-effectors of the kinematic model, a plurality of virtual end-effectors corresponding to the target positions, wherein each of the virtual end-effectors is generated at a point on the model skin of the kinematic model closest to an associated target position of the target positions;
generating an inverse kinematics problem comprising a Jacobian matrix based on the initial kinematic model parameters, the target positions, and the virtual end-effectors;
determining a change in the kinematic model parameters based on the inverse kinematics problem;
repeating the selecting the p

Assignees

• Intel Corp

Inventors

• Fleishman Shahar
• Kliger Mark
• Lerner Alon

Classifications

• G06T2207/30241

  Trajectory · CPC title

• G06T7/11

  Region-based segmentation · CPC title

• G06T7/75

  involving models · CPC title

• G06T2207/10004

  Still image; Photographic image · CPC title

• G06T17/10

  Constructive solid geometry [CSG] using solid primitives, e.g. cylinders, cubes · CPC title