An orientation designated in advance as an orientation that a grip unit is to take to grip an object having a shape of rotational symmetry with respect to at least one axis is acquired as a reference orientation. The relative position and orientation of the object and grip unit when the grip unit grips the object is acquired as a taught position and orientation. The position and orientation of the object is recognized from an image, and an initial position and initial orientation in which the grip unit grips the object are derived based on the recognized and taught positions and orientations. A grip orientation to grip the object is decided based on the reference and the initial orientation, and a grip position and orientation in which the grip unit grips the object is decided based on the grip orientation and the initial position.