IP Library Granted Patent US 10,189,162
Granted Patent B2
US 10,189,162 · App. 13/788,237 · Granted Jan 29, 2019

Model generation apparatus, information processing apparatus, model generation method, and information processing method

Inventors: Keisuke Tateno (Kawasaki, JP); Hisayoshi Furihata (Tokyo, JP); Shinji Uchiyama (Yokohama, JP)
Assignee: Canon Kabushiki Kaisha
B25J9/1697G06K9/00214G06T7/75G06T17/00G06T2207/10004G06T2207/10012G06T2207/10028
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 10,189,162
App. No.
13/788,237
Granted
Jan 29, 2019
Kind
B2
Abstract

A three-dimensional shape model of a target object is input. A position and orientation of at least one image sensing device used to capture an image of the target object is set so as to virtually set a relative position and orientation between the target object and the image sensing device for the three-dimensional shape model of the target object. At least one position and orientation of the image sensing device is selected. Geometric features are grouped based on a relationship between an image to be obtained at the selected position and orientation of the image sensing device and the geometric features of the three-dimensional shape model.

Claims (27)

1. A model generation method for execution by a model generation apparatus, the model generation method comprising:

inputting a three-dimensional shape model of a target object including (1) plane geometric features and (2) line geometric features;

setting a plurality of positions and orientations of a viewpoint, the viewpoint being used to observe the three-dimensional shape model of the target object, and sequentially selecting one of the plurality of positions and orientations of the viewpoint;

calculating, for the selected one of the plurality of positions and orientations of the viewpoint set in the setting and for the plane geometric features and the line geometric features, respectively, a degree of change of a relationship between image features of an image to be observed at the viewpoint and each of the plane geometric features and the line geometric features of the three-dimensional shape model when the selected position and orientation of the viewpoint or a position and orientation of the target object is changed; and

generating a hierarchical cluster of plane geometric features and a hierarchical cluster of line geometric features, separately, for the selected one of the plurality of positions and orientations of the viewpoint, by setting each of the plane geometric features and each of the line geometric features as an initial cluster in a lowest layer, and sequentially merging a pair of clusters in a current layer to generate a cluster in a higher layer based on similarity of the degree of change of the relationship calculated for the pair of clusters.

2. The model generation method according to claim 1 , further comprising:

selecting, from the plurality of positions and orientations of the viewpoint set in the setting, at least one position and orientation of an image sensing device,

wherein, in the generating, the plane geometric features are grouped based on the degree of change of the relationship between image features of the image to be observed by an image sensing device and the plane geometric features of the three-dimensional shape model when (a) the position and orientation of the image sensing device selected by the selecting or (b) the position and orientation of the target object is changed.

3. The model generation method according to claim 1 , wherein, in the generating, Jacobians of the plane geometric features of the three-dimensional shape model of the target object are calculated as a value which indicates a degree of change of a relationship between image features of an image to be observed at the viewpoint and the plane geometric features of the three-dimensional shape model,

wherein the plane geometric features of the three-dimensional shape model of the target object are to be grouped in a mapping relationship of the three-dimensional shape model of the target object onto the image to be observed at the viewpoint, and

wherein, in the generating, the geometric features are grouped based on similarity of the calculated Jacobians.

4. The model generation method according to claim 1 , wherein, in the setting, the three-dimensional shape model of the target object is further enclosed by a sphere, latitudes and longitudes are assigned to the sphere, and virtual positions and orientations of an image sensing device are set using the assigned latitudes and longitudes.

5. The model generation method according to claim 1 , wherein the image to be observed comprises one or both of a two-dimensional image and a range image.

6. The model generation method according to claim 1 , further comprising:

saving, as a viewpoint-dependent model, a set of a result of the generating and the viewpoint for observing the three-dimensional shape model of the target object.

7. A non-transitory computer-readable storage medium storing a computer program which upon execution enables a computer to carry out the steps of:

inputting a three-dimensional shape model of a target object including (1) plane geometric features and (2) line geometric features;

setting a plurality of positions and orientations of a viewpoint, the viewpoint being used to observe the three-dimensional shape model of the target object, and sequentially selecting one of the plurality of positions and orientations of the viewpoint;

calculating, for the selected one of the plurality of positions and orientations of the viewpoint set in the setting and for the plane geometric features and the line geometric features, respectively, a degree of change of a relationship between image features of an image to be observed at the viewpoint and each of the plane geometric features and the line geometric features of the three-dimensional shape model when the selected position and orientation of the viewpoint or a position and orientation of the target object is changed; and

generating a hierarchical cluster of plane geometric features and a hierarchical cluster of line geometric features, separately, for the selected one of the plurality of positions and orientations of the viewpoint, by setting each of the plane geometric features and each of the line geometric features as an initial cluster in a lowest layer, and sequentially merging a pair of clusters in a current layer to generate a cluster in a higher layer based on similarity of the degree of change of the relationship calculated for the pair of clusters.

8. A model generation apparatus comprising:

one or more processors; and

a memory coupled to the one or more processors and having instructions stored thereon, wherein the instructions, when executed by the one or more processors, cause the model generation apparatus to:

(1) input a three-dimensional shape model of a target object including (1) plane geometric features and (2) line geometric features;

(2) set a plurality of positions and orientations of a viewpoint, the viewpoint being used to observe the three-dimensional shape model of the target object, and sequentially selecting one of the plurality of positions and orientations of the viewpoint;

(3) calculate, for the selected one of the plurality of positions and orientations of the viewpoint set in the setting and for the plane geometric features and the line geometric features, respectively, a degree of change of a relationship between image features of an image to be observed at the viewpoint and each of the plane geometric features and the line geometric features of the three-dimensional shape model when the selected position and orientation of the viewpoint or a position and orientation of the target object is changed; and

generate a hierarchical cluster of plane geometric features and a hierarchical cluster of line geometric features, separately, for the selected one of the plurality of positions and orientations of the viewpoint, by setting each of the plane geometric features and each of the line geometric features as an initial cluster in a lowest layer, and sequentially merging a pair of clusters in a current layer to generate a cluster in a higher layer based on similarity of the degree of change of the relationship calculated for the pair of clusters.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jun 12, 2013
From: TATENO, KEISUKE; FURIHATA, HISAYOSHI; UCHIYAMA, SHINJI
To: CANON KABUSHIKI KAISHA
Reel/Frame 030592/0215 →
Priority Claims (2)
JP 2012-056489 · Mar 13, 2012 · national
JP 2012-167647 · Jul 27, 2012 · national
Continuity (1)
Related Publication 20130245828A1 · Sep 19, 2013
Cited By (1)
US 12,276,522