Legged mobile robot
First Claim
Patent Images
1. A legged mobile robot comprising:
- a base body;
a plurality of legs operatively connected to said base body;
a vision sensor mounted on said base body;
image input means for capturing image data via said vision sensor;
positional relationship calculating means for calculating the positional relationship between of an obstacle present in the course of robot walking and one of said legs from the image data captured by said image input means, so that the legged mobile robot can walk while recognizing the positional relationship between the obstacle and the leg as calculated by said positional relationship calculating means;
said positional relationship calculating means comprising means for calculating positional relationship between the obstacle and the leg from image data containing both images of a portion of the leg and the obstacle as captured by said image input means.
1 Assignment
0 Petitions
Accused Products
Abstract
A legged mobile robot for accurately calculating the positional relationship between a leg and an obstacle to the walking of the robot has a base body and a plurality of legs operatively connected to the base body. A vision sensor mounted on the base body captures image data containing both an image of a portion of a foot of one of the legs and an image of an obstacle in the course of robot walking. The legged mobile robot calculates the positional relationship between the leg and the obstacle from the captured image data, and walks while recognizing the positional relationship as calculated between the obstacle and the foot.
155 Citations
8 Claims
-
1. A legged mobile robot comprising:
-
a base body;
a plurality of legs operatively connected to said base body;
a vision sensor mounted on said base body;
image input means for capturing image data via said vision sensor;
positional relationship calculating means for calculating the positional relationship between of an obstacle present in the course of robot walking and one of said legs from the image data captured by said image input means, so that the legged mobile robot can walk while recognizing the positional relationship between the obstacle and the leg as calculated by said positional relationship calculating means;
said positional relationship calculating means comprising means for calculating positional relationship between the obstacle and the leg from image data containing both images of a portion of the leg and the obstacle as captured by said image input means. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8)
positional relationship recognizing means for recognizing the positional relationship between a visual field of said vision sensor and said leg, said image input means comprising means for capturing the image data via said vision sensor when both said portion of the leg and the obstacle enter the visual field of said vision sensor as determined by said positional relationship recognizing means.
-
-
3. A legged mobile robot according to claim 1, further comprising:
walking operation modifying means for modifying details of subsequent walking operation of the legged mobile robot depending on the positional relationship between said leg and the obstacle as calculated by said positional relationship calculating means.
-
4. A legged mobile robot according to claim 3, wherein said positional relationship calculating means comprises means for calculating the positional relationship between said leg and the obstacle at least once after said leg is lifted and until said leg is landed during walking operation of the legged mobile robot.
-
5. A legged mobile robot according to claim 1, wherein said image input means comprises means for holding topographic information in advance and means for capturing the image data via said vision sensor when an area in which the obstacle is expected to enter the visual field of said vision sensor is reached based on the topographic information during walking operation of the legged mobile robot.
-
6. A legged mobile robot according to claim 1, wherein said leg has a mark on a portion thereof which can enter the visual field of said vision sensor, said positional relationship calculating means comprising means for extracting an edge of said mark from said image data thereby to calculate the position of said leg.
-
7. A legged mobile robot according to claim 1, further comprising:
-
an auxiliary vision sensor having a visual field wider than a visual field of said vision sensor;
said image input means comprising means for entering topographic image data in the course of robot walking via said auxiliary vision sensor, recognizing the position of the obstacle from said topographic image data, and capturing the image data via said vision sensor when an area in which the obstacle enters the visual field of said vision sensor is reached.
-
-
8. A legged mobile robot according to claim 1, wherein said leg has a foot for contacting a floor, further comprising:
-
a force sensor for detecting an acting force and a moment in a predetermined direction based on a resistive force applied to said foot from the floor; and
contacted state recognizing means for recognizing a direction and/or position in which said foot contacts a ridge of the obstacle, based on output data from said force sensor;
wherein if said image input means fails to capture the image data via said vision sensor, the legged mobile robot walks while recognizing the positional relationship between the obstacle and the leg from the direction and/or position in which said foot contacts the ridge of the obstacle, as recognized by said contacted state recognizing means.
-
Specification