Abstract: Disclosed herein is a computer-readable medium and method of a mobile platform detecting and tracking dynamic objects in an environment having the dynamic objects. The mobile platform acquires a three-dimensional (3D) image using a time-of-flight (TOF) sensor, removes a floor plane from the acquired 3D image using a random sample consensus (RANSAC) algorithm, and individually separates objects from the 3D image. Movement of the respective separated objects is estimated using a joint probability data association filter (JPDAF).

Abstract: Disclosed is a humanoid robot apparatus, method and computer-readable medium thereof related to lifting and holding a heavy object having a weight unknown to the robot, by measuring an external force acting on the robot. Linear momentum and rotational momentum are compensated for stepwise according to the degree of stability of the robot which is determined based on the measured external force. Accordingly, the robot stably lifts and holds the object without losing its balance.

Abstract: Disclosed is an apparatus and method adjusting motion of each joint of a robot to compensate for friction force of each joint such that the sole of the foot of the robot clings to the ground. The motion of each joint is adjusted as if gravity acts on each joint of the robot in a direction opposite to gravity and the robot is held in an erect state. Therefore, the robot can stand while keeping its balance without falling.

Abstract: Disclosed are an apparatus, a method and a computer-readable medium controlling whole-body operation of a humanoid robot. The humanoid robot recognizes a motion control code using binary data mapped according to a motion command represented by a language understood by a human to implement a whole-body operation. Since a control mode corresponding to a task space control and a control mode corresponding to a joint space control are used together to describe whole-body motion, the whole-body operation more similar to a human action may be easily implemented.

Abstract: Disclosed herein are a method and apparatus for extracting feature points using hierarchical image segmentation and an image based localization method using the extracted feature points. An image is segmented using an affinity degree obtained using information observed during position estimation, new feature points are extracted from segmented areas in which registered feature points are not included, and position estimation is performed based on the new feature points. Accordingly, stable and reliable localization may be performed.

Abstract: A suitable waypoint is selected using a goal score, a section from a start point to a goal point through the waypoint is divided into a plurality of sections based on the waypoint with a solution of inverse kinematics, and trees are simultaneously expanded in the sections using a Best First Search & Rapidly Random Tree (BF-RRT) algorithm so as to generate a path. By this configuration, a probability of local minima occurring is decreased compared with the case where the waypoint is randomly selected. In addition, since the trees are simultaneously expanded in the sections each having the waypoint with a solution of inverse kinematics, the solution may be rapidly obtained. A time consumed to search for an optimal motion path may be shortened and path plan performance may be improved.

Abstract: If a manipulator of a robot falls in local minima when expanding a node to generate a path, the manipulator may efficiently escape from local minima by any one of a random escaping method and a goal function changing method or a combination thereof to generate the path. When the solution of inverse kinematics is not obtained due to local minima or when the solution of inverse kinematics is not obtained due to an inaccurate goal function, an optimal motion path to avoid an obstacle may be efficiently searched for. The speed to obtain the solution may be increased and thus the time consumed to search for the optimal motion path may be shortened.

Abstract: Disclosed herein is a path planning apparatus and method of a robot, in which a path, along which the robot accesses an object to grasp the object, is planned. The path planning method includes judging whether or not a robot hand of a robot collides with an obstacle when the robot hand moves along one access path candidate selected from plural access path candidates along which the robot hand accesses an object to grasp the object, calculating an access score of the selected access path candidate when the robot hand does not collide with the obstacle, and determining an access path plan using the access score of the selected access path candidate.

Abstract: Disclosed herein are a path planning apparatus and a method for a robot to plan an optimal path along which a manipulator of a robot moves to a goal point from a start point. An obstacle within a prescribed angle on a straight line connecting a start point and a goal point is recognized as a middle point in a configuration space and arbitrary points separated from the middle point by a prescribed distance are selected. Among the selected points, arbitrary points which can directly connect the start point and the goal point without passing the obstacle are selected as waypoints to map a new middle node. A path is extended via the middle node and extension of a tree in a wrong direction is minimized so that the manipulator is not struck at local minima without depending greatly on a goal score, thereby improving the performance of path planning and rapidly searching for a path.

Abstract: An omni-directional stereo camera and a control method thereof. The omni-directional stereo camera includes two or more omni-directional cameras, and a supporting member installed within a shooting range between the omni-directional cameras to interconnect the omni-directional cameras and including compensation patterns formed at the surfaces.

Abstract: Disclosed herein are a feature point used to localize an image-based robot and build a map of the robot and a method of extracting and matching an image patch of a three-dimensional (3D) image, which is used as the feature point. It is possible to extract the image patch converted into the reference image using the position information of the robot and the 3D position information of the feature point. Also, it is possible to obtain the 3D surface information with the brightness values of the image patches to obtain the match value with the minimum error by a 3D surface matching method of matching the 3D surface information of the image patches converted into the reference image through the ICP algorithm.

Abstract: Provided is a method of using a rotational movement amount of a mobile device including obtaining images at two different points on a path along which the mobile device moves, searching for matching points with respect to the obtained images and obtaining an image coordinate value of each of the matching points, sensing linear movement of the mobile device and obtaining a linear movement amount using a result of sensing, and obtaining the rotational movement amount using the image coordinate values and the linear movement amount.

Abstract: Disclosed herein is a method of building a map of a mobile platform moving in a dynamic environment and detecting an object using a 3D camera sensor, e.g., an IR TOF camera sensor, for localization. A localization technology to separate and map a dynamic object and a static object is applied to a mobile platform, such as an unmanned vehicle or a mobile robot. Consequently, the present method is capable of accurately building map information based on the static object in a dynamic environment having a large number of dynamic objects and achieving a dynamic object avoidance or chasing function using position information acquired to build the map.

Abstract: Disclosed is a method of defining control angles to use a limit cycle in order to balance a biped walking robot on a three-dimensional space. In order to balance an FSM-based biped walking robot right and left on a three-dimensional space, limit cycle control angles to balance the robot according to states of the FSM-based robot are set on the three-dimensional space, and the limit cycle control angles on the three-dimensional space are controlled using a sinusoidal function to allow relations between the control angles and control angular velocities to form a stable closed loop within a limit cycle.

Abstract: Disclosed is a method of generating a hip trajectory of a biped walking robot to allow the robot to stably walk on a two-dimensional space without falling down. An angular velocity of a hip of a swinging leg is obtained by measuring the angle/angular velocity of an ankle pitch joint part of a supporting leg in real time when the robot walks on the two-dimensional space, and desired trajectories of the ankle and the hip are generated based on the angular velocity of the ankle of the supporting leg and the angular velocity of the hip of the swinging leg.

Abstract: An finite state machine (FSM)-based biped walking robot, to which a limit cycle is applied to balance the robot right and left on a two-dimensional space, and a method of controlling balance of the robot. In order to balance an FSM-based biped walking robot right and left on a two-dimensional space, control angles to balance the robot according to states of the FSM-based biped walking robot are set, and the control angles are controlled using a sinusoidal function to allow relations between the control angles and control angular velocities to form a stable closed loop within a limit cycle, thereby allowing the biped walking robot to balance itself while changing its supporting foot and thus to safely walk without falling down.

Abstract: A finite state machine (FSM)-based biped robot, to which a limit cycle is applied to balance the robot right and left on a two-dimensional space, and a method of controlling balance of the robot. In order to balance an FSM-based biped robot right and left on a two-dimensional space, control angles to balance the robot according to states of the FSM-based biped robot are set. The range of the control angles is restricted to reduce the maximum right and left moving distance of the biped robot and thus to reduce the maximum right and left moving velocity of the biped robot, thereby reducing the sum total of the moments of the biped robot and thus allowing the ankles of the biped robot to balance the biped robot to be controlled, and causing the soles of the feet of the biped robot to parallel contact the ground.

Abstract: Disclosed are a robot arm having high back-drivability to interact with a human being, which is safely stopped, and a method of controlling the robot arm. When the operation of the robot arm having high back-drivability to interact with a human being is converted into a safety mode due to the occurrence of an emergency, in which the operation of the robot arm is stopped, only a torque having a degree to compensate for gravity applied to the robot arm, i.e., only a torque to maintain the kinematical position of the current state of the robot arm without falling of the robot arm due to gravity, is outputted to control the robot arm as if the robot arm is in a weightless state, thus being capable of safely stopping the robot arm. Thereby, a user can easily move the robot arm and thus safely escape from the robot arm, and it is possible to prevent the falling of the robot arm due to gravity and thus to protect devices of the robot arm and peripheral environment around the robot arm.

Abstract: There is provided a robot, which compensates for the angle of a joint, bent due to external force, using a redundant degree of freedom of the robot, and a method of controlling the same. The method includes sensing a bending angle of a joint due to external force or a torque applied to the joint, comparing the bending angle of the joint or the torque applied to the joint with an allowable safety reference value; and adjusting the bending angle of a higher-level joint of the joint using a redundant degree of freedom of the robot, when the bending angle or the torque reaches the allowable safety reference value, thereby compensating for the excessively bending angle of the joint and thus being capable of safely controlling the robot.

Abstract: A method and apparatus to build a 3-dimensional grid map and a method and apparatus to control an automatic traveling apparatus using the same. In building a 3-dimensional map to discern a current location and a peripheral environment of an unmanned vehicle or a mobile robot, 2-dimensional localization and 3-dimensional image restoration are appropriately used to accurately build the 3-dimensional grid map more rapidly.