An image processing device includes: an augmented-reality marker detecting unit that detects, from a captured image, an augmented reality marker indicating that a virtual object is to be displayed; an operation-marker detecting unit that detects, from the captured image, an area of an operation marker that is provided so as to correspond to the augmented reality marker; an operation detecting unit that detects, from the area of the operation marker, a user's operation on the operation marker; a virtual-object control unit that deforms a virtual object corresponding to the augmented reality marker in accordance with the user's operation; and a drawing unit that draws the deformed virtual object in the captured image in a superimposed manner.