Object detecting method, object detecting device, and robot system
View Patent ↗An object detecting method includes imaging a plurality of target objects with an imaging section and acquiring a first image, recognizing an object position/posture of one of the plurality of target objects based on the first image, counting the number of successfully recognized object positions/postures of the target object, outputting, based on the object position/posture of the target object, a signal for causing a holding section to hold the target object, calculating, as a task evaluation value, a result about whether the target object was successfully held, updating, based on an evaluation indicator including the number of successfully recognized object positions/postures and the task evaluation value, a model for estimating the evaluation indicator from an imaging position/posture of the imaging section and determining an updated imaging position/posture, acquiring a second image in the updated imaging position/posture, and recognizing the object position/posture of the target object based on the second image.
1. An object detecting method for detecting an object position/posture of a target object, the object detecting method comprising:
imaging a plurality of target objects with an imaging section and acquiring a first image;
recognizing the object position/posture of the target object based on the first image;
counting, as a number of recognized object positions/postures, a number of successfully recognized object positions/postures of the target object;
outputting, based on the object position/posture of the target object, a signal for causing a holding section to hold the target object;
calculating, as a task evaluation value, a result about whether the target object was successfully held;
updating, based on an evaluation indicator including the number of recognized object positions/postures and the task evaluation value, a model for estimating the evaluation indicator from an imaging position/posture of the imaging section and determining an updated imaging position/posture of the imaging section based on the model after the update;
imaging the plurality of target objects in the updated imaging position/posture and acquiring a second image; and
recognizing the object position/posture of the target object based on the second image.
2. The object detecting method according to claim 1 , wherein the first image is a two-dimensional image or a depth image.
3. The object detecting method according to claim 1 , wherein the model is a model of Bayesian inference into which a concept of a forgetting rate is introduced.
4. The object detecting method according to claim 1 , wherein the task evaluation value includes a result obtained by causing the holding section to hold the target object and perform work using the target object.
5. An object detecting device that detects an object position/posture of a target object, the object detecting device comprising:
an imaging control section configured to capture, with an imaging section, a first image including a plurality of target objects and acquire the first image;
an object-position/posture calculating section configured to recognize the object position/posture of the target object based on the first image;
a recognition evaluating section configured to count, as a number of recognized object positions/postures, a number of successfully recognized object positions/postures of the target object;
a holding-position/posture calculating section configured to calculate, based on the object position/posture of the target object, a holding position/posture of a holding section that holds the target object and output a control signal for causing the holding section to hold the target object in the holding position/posture;
a task evaluating section configured to acquire a result about whether the target object was successfully held by the holding section and calculate a task evaluation value; and
an imaging-position/posture determining section configured to update, based on an evaluation indicator including the number of recognized object positions/postures and the task evaluation value, a model for estimating the evaluation indicator from an imaging position/posture of the imaging section and determine an updated imaging position/posture of the imaging section based on the model after the update, wherein
the imaging control section causes the imaging section to capture a second image in the updated imaging position/posture and acquires the second image, and
the object-position/posture calculating section recognizes the object position/posture of the target object based on the second image.
6. The object detecting device according to claim 5 , further comprising a display section configured to display at least one of the number of recognized object positions/postures, the task evaluation value, and the evaluation indicator.
7. A robot system comprising:
a robot including a robot arm;
an imaging section set in the robot arm;
the object detecting device according to claim 5 ; and
a robot control device configured to control driving of the robot based on a detection result of the object detecting device.
8. The robot system according to claim 7 , further comprising an automated guided vehicle mounted with the robot.
9. The robot system according to claim 7 , wherein the imaging-position/posture determining section initializes the model when a position of the robot changes.