Operating system, control device, and computer program product

1. An operating system comprising:

a movable member;

an actuator that moves the movable member;

a storage; and

one or more hardware processors configured to implement:

a target surface detection unit that detects a target surface of a target object from a depth image obtained by a depth sensor, wherein coordinates in the depth image are stored in the storage for at least one point in the detected target surface that is detected at a first time;

a position calculation unit that calculates a first position for the detected target surface;

a direction calculation unit that calculates a first direction for the detected target surface; and

a movement control unit that controls the actuator so as to reduce a positional deviation between a second position fixed with respect to the movable member and the first position and to reduce a directional deviation between a second direction fixed with respect to the movable member and the first direction, wherein

the target surface detection unit sets a specific range that is narrower than a whole region of the depth image based on a stored coordinate, and detects a first region of the detected target surface within the specific range and detects a second region linked to the first region as the target surface outside the specific range at a second time after the first time.

2. The operating system according to claim 1 , wherein the direction calculation unit calculates the first direction as a normal direction of an approximate plane of the detected target surface.

3. The operating system according to claim 1 , wherein the direction calculation unit calculates, for each pixel in the detected target surface, a normal direction from a three-dimensional position of a proximity region including the each pixel and a plurality of pixels adjacent to the pixel, and calculates the first direction as a representative value of the normal direction in the plurality of pixels in the detected target surface.

4. The operating system according to claim 3 , wherein the direction calculation unit calculates the normal direction from an outer product of two vectors along the proximity region.

5. The operating system according to claim 1 ,

wherein

the one or more hardware processors are configured to further implement an angular deviation calculation unit that calculates an angular deviation from a given posture of the movable member in rotation around an axis along the first direction of the detected target surface, and

the movement control unit controls the actuator that moves the movable member so as to further reduce the angular deviation.

6. The operating system according to claim 5 , wherein the one or more hardware processors are configured to further implement a target object determination unit that includes a network learned by learning data generated based on imaging information that is transformed into three-dimensional information and a physical feature of the movable member, and determines the target object corresponding to the acquired imaging information.

7. A control device comprising:

one or more hardware processors configured to implement:

a target surface detection unit that detects a target surface of a target object from a depth image obtained by a depth sensor, wherein coordinates in the depth image are stored in a storage for at least one point in the detected target surface that is detected at a first time,

a position calculation unit that calculates a first position for the detected target surface,

a direction calculation unit that calculates a first direction for the detected target surface, and

a movement control unit that controls an actuator that moves a movable member so as to reduce a positional deviation between a second position fixed with respect to the movable member and the first position and to reduce a directional deviation between a second direction fixed with respect to the movable member and the first direction, wherein

the target surface detection unit sets a specific range that is narrower than a whole region of the depth image based on a stored coordinate, and detects a first region of the detected target surface within the specific range and detects a second region linked to the first region as the target surface outside the specific range at a second time after the first time.

8. The control device according to claim 7 , wherein the one or more hardware processors are configured to further implement a target object determination unit that includes a network learned by learning data generated based OD imaging information that is transformed into three-dimensional information and a physical feature of the movable member, and determines the target object corresponding to the acquired imaging information.

9. A computer program product having a non-transitory computer readable medium including instructions, wherein the instructions, when executed by a computer, cause the computer to function as the target surface detection unit, the position calculation unit, the direction calculation unit, and the movement control unit of the operating system according to claim 1 .

10. A method comprising:

detecting, by one or more hardware processors, a target surface of a target object from a depth image obtained by a depth sensor, wherein coordinates in the depth image are stored in a storage for at least one point in the detected target surface that is detected at a first time;

calculating, by the one or more hardware processors, a first position for the detected target surface;

calculating, by the one or more hardware processors, a first direction for the detected target surface; and

controlling, by the one or more hardware processors, an actuator that moves a movable member so as to reduce a positional deviation between a second position fixed with respect to the movable member and the first position and to reduce a directional deviation between a second direction fixed with respect to the movable member and the first direction; and

setting, by the one or more hardware processors, a specific range that is narrower than a whole region of the depth image based on a stored coordinate, and detecting a first region of the detected target surface within the specific range and detects a second region linked to the first region as the target surface outside the specific range at a second time after the first time.