A contact region between an object and a holding member is acquired based on the image obtained by imaging the object in a held state in which the object is held by the holding member. The three-dimensional shape of the object in a held state is estimated based on the contact region.