Image compression by pointwise prototype correspondence using shape and texture information

• US 5,844,573 A
• Filed: 08/02/1996
• Issued: 12/01/1998
• Est. Priority Date: 06/07/1995
• Status: Expired due to Fees

• Current Assignee: Massachusetts Institute of Technology
• Original Assignee: Massachusetts Institute of Technology
• Inventors: Jones, Michael, Vetter, Thomas, Poggio, Tomaso, Beymer, David
• Primary Examiner(s): Jankus, Almis R.
• Application Number:
  US08/691,550
• Time in Patent Office: 851 Days
• Field of Search: 345/441, 345/433-439, 382/118, 382/201, 382/224, 348/2, 455/2
• US Class Current: 345/441
• CPC Class Codes: G06T 11/00 2D [Two Dimensional] image generation
• CPC Class Codes: G06T 9/001 Model-based coding, e.g. wire frame

Abstract

An image processing system (100) analyzes novel images in one of three modes. In a first mode, the system learns a model based on vectorized prototype images, and uses the model to characterize the shape of a novel image as a pixel-by-pixel linear combination of the prototype shapes. In a second mode, both shape and texture vectors are used to characterize the novel image as a linear combination of the prototypes. In a third mode, shape information is modeled by a linear combination of prototypes and texture information is modeled by an optical flow, and processing of shape and texture information is interleaved. Novel images are synthesized by analysis of input images using transformed prototypes. Image compression is achieved by storing or transmitting parameter vectors relating the novel or applied images to the prototypes.

Claims

1. A method of compressing an applied physical image, comprising:

(a) providing a plurality of prototype images;

(b) determining pointwise prototype correspondences among the plurality of prototype images;

(c) creating a model responsive to the pointwise prototype correspondences, the model being a parameterized pointwise weighted linear combination of vectors associated with the prototypes, the vectors being responsive to texture;

(d) filtering and normalizing the applied physical image;

(e) determining parameters to minimize a pointwise distance metric from the applied physical image to the model, the pointwise distance metric being responsive to said texture; and

(f) representing the new physical image based on the parameters.

2. A method as in claim 1, wherein the vectors associated with the prototypes are responsive to shape, and the pointwise distance metric is responsive to said shape.

3. A method of compressing an applied physical image having an applied image shape and an applied image texture, the process comprising:

(a) providing a plurality of prototype images;

(b) determining pointwise prototype correspondences among the plurality of prototype images, thereby associating to each prototype image a prototype shape and a prototype texture;

(c) creating a model as a linear combination of geometrically normalized textures derived from the prototypes, the model having a model shape and a model texture;

(d) parameterizing the applied image shape responsive to optical flow transformation of the applied image shape to the model shape, thereby producing a shape parameter vector set;

(e) parameterizing the applied image texture responsive to pointwise combinations of the textures of the prototypes, thereby producing a texture parameter vector set; and

(f) representing the applied physical image based on the shape parameter vector set and the texture parameter vector set, said representing including representing the applied physical image by eigenvectors derived from the shape parameter vector set and the texture parameter vector set.

4. An image processing system for compressing an applied image of a physical object using a plurality of prototypes, the system comprising:

an imaging device for producing a signal responsive to the applied image;

a prototype storage device, the prototype storage device containing the plurality of prototypes;

a texture engine;

a shape engine operatively connected to the imaging device and to the prototype storage device, the shape engine having a synthesis mode of determining pointwise prototype correspondence among the plurality of prototypes and creating a model responsive to the pointwise prototype correspondences, the shape engine further having an operating mode of filtering and normalizing the signal to produce a preprocessed signal, determining model parameters to minimize a pointwise distance metric from the applied image to the model responsive to the preprocessed signal, establishing pointwise correspondences responsive to the plurality of prototypes and the preprocessed signal, the shape engine cooperating with the texture engine to produce the model parameters; and

an output engine operatively connected to the shape engine, the output engine producing output representing a compressed version of the applied image, the compressed version being responsive to the model parameters and the pointwise correspondences.

5. An image processing system for compressing an applied image of a physical object using a plurality of prototypes, the system comprising:

an imaging device for producing a signal responsive to the applied image;

a prototype storage device, the prototype storage device containing the plurality of prototypes;

a texture engine for processing texture information responsive to the signal;

a shape engine operatively connected to the imaging device, to the prototype storage device, and to the texture engine, the shape engine and the texture engine cooperating to provide a synthesis mode of determining pointwise prototype correspondences among the plurality of prototypes and creating a model responsive to linear combinations of geometrically normalized textures derived from the prototypes, the shape engine and the texture engine further cooperating to provide an operating mode of filtering and normalizing the signal to produce a preprocessed signal, and iteratively parameterizing image shape and texture responsive to the plurality of prototypes and the preprocessed signal; and

an output engine operatively connected to the shape engine, the output engine producing output representing a compressed version of the applied image, the compressed version being responsive to the model parameters and the pointwise correspondences.