Strict recognizing apparatus using observation point

• US 5,864,779 A
• Filed: 12/03/1996
• Issued: 01/26/1999
• Est. Priority Date: 02/29/1996
• Status: Expired due to Term

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Fujimoto, Katsuhito
• Primary Examiner(s): Arana, Louis M.
• Assistant Examiner(s): ASSOUAD, PATRICK J
• Application Number:
  US08/758,310
• Time in Patent Office: 784 Days
• Field of Search: 364/556, 364/559, 364/560, 364/561, 382/155, 382/168, 382/170, 382/171, 382/172, 382/181, 382/190, 382/195, 382/281, 382/286, 382/288, 382/291, 382/307, 358/462, 358/467, 395/93, 345/418, 348/129, 348/130, 348/135, 348/161, 702/179, 702/150, 702/180
• US Class Current: 702/179
• CPC Class Codes: G06V 10/48 by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation

Abstract

A recognizing apparatus for estimating an environment description parameter, and substantially reducing a memory capacity required to represent a parameter space. The recognizing apparatus stores a restricted parameter space as a parameter space restricted by predetermined observation point information and based on an object model as shape information. The apparatus votes for a parameter subset consistent with the object model for each characteristic point obtained through environment observation in a restricted parameter space. The apparatus then outputs an estimated value for a environment description parameter according to the result of the voting for the restricted parameter space.

Claims

1. A strict recognizing apparatus implemented on a computer using an observation point in estimating a parameter describing an environment using a set of characteristic points obtained by observing the environment and an object model implying shape information of an object assumed to exist in the environment, comprising:

restricted parameter space storing means for storing a restricted parameter space as a space in the parameter generated based on the object model and restricted by predetermined observation point information corresponding to the shape information about the object model;

voting means for voting for a subset of parameters matching the object model in each of the characteristic points in the restricted parameter space; and

parameter estimating means for outputting an estimated value of the parameter describing the environment corresponding to a voting result for a voting space in the restricted parameter space after taking votes for all of the characteristic points.

2. The strict recognizing apparatus according to claim 1, further comprising:

object model storing means for storing the shape information for the object model and outputting the shape information to said restricted parameter space storing means and said voting means.

3. The strict recognizing apparatus according to either claim 1 or 2,

wherein said shape information about the object model is a set of position coordinates of outline points of a two-dimensional object;

and

wherein said observation point information refers to an optional position coordinate of the outline points.

4. The strict recognizing apparatus according to claim 1 or 2,

wherein said shape information about the object model consists of a set of position-and-postures of the outline normals of a two-dimensional object;

and

wherein said observation point information refers to an optional position-and-posture of the outline normals.

5. The strict recognizing apparatus according to claim 1 or 2,

wherein said shape information about the object model is a set of position coordinates of surface points of a three-dimensional object;

and

wherein said observation point information refers to an optional position coordinate of the surface points.

6. The strict recognizing apparatus according to claim 1 or 2,

wherein said shape information about the object model is a set of position-and-postures of normals on a surface of a three-dimensional object;

and

wherein said observation point information refers to an optional position-and-posture of the normals on the surface.

7. The strict recognizing apparatus according to claim 1 or 2,

wherein said estimated value of the parameter output by said parameter estimating means consists of an estimated value of position coordinates of a center of gravity of the object.

8. The strict recognizing apparatus according to claim 1 or 2,

wherein said estimated value of the parameter output by said parameter estimating means consists of an estimated value of a position-and-posture of the object as a transformation parameter from an object model coordinate system for the object into a work coordinate system set in the environment.

9. A method for estimating a parameter describing an environment using a set of characteristic points obtained by observing the environment and an object model implying shape information of an object assumed to exist in the environment, including a computer-readable storage medium and a computer, comprising the steps of:

storing a restricted parameter space as a space in the parameter generated based on the object model and restricted by predetermined observation point information corresponding to the shape information about the object model;

storing a voting space in which a vote is taken for a subset of parameters as matching the object model in each of the characteristic points in the restricted parameter space; and

outputting an estimated value of the parameter describing the environment corresponding to a voting result for a voting space in the restricted parameter space after taking votes for all of the characteristic points.

10. A strict recognizing method implemented on a computer using an observation point in estimating a parameter describing an environment using a set of characteristic points obtained by observing the environment and an object model implying shape information of an object assumed to exist in the environment, comprising the steps of:

storing a restricted parameter space as a space in the parameter generated based on the object model and restricted by predetermined observation point information corresponding to the shape information about the object model;

voting for a subset of parameters as matching the object model in each of the characteristic points in the restricted parameter space; and

outputting an estimated value of the parameter describing the environment corresponding to a voting result for a voting space in the restricted parameter space after taking votes for all of the characteristic points.