Model based image editing and correction

A system and method is described for automatically editing digital images. An image to be edited is selected and registered with a model that is an abstract of features and attributes common to a category of images to which the selected image belongs. In the illustrated embodiment, a human face is used as an example of an image to be edited and the matching model is a representation of features and attributes common to all human faces. Many other images of objects may also be edited using the techniques of the invention, as long as a general class of objects to which an image belongs is amenable to being represented by a model comprising a set of common features or attributes. A library of different models is available to the user for selecting the model matching the image to be edited. The library includes different models of the same object in different orientations and may also include models of different objects. After the appropriate model is selected, the user is prompted to register the details of the selected image with the generalized features and attributes of the selected model. Editing processes or tools, which are linked to the model, are then automatically appled to the image in order to perform a desired editing function. In addition to applying editing techniques to existing images in a larger image, the invention enables a user to paste and blend images into another, larger image-e.g., create a morphing effect by pasting one persons head over the head of another in an original image.