Apparatus and method for picture representation by data compression
AssigneeIGP, Research and Development Ltd.
InventorsYosef Yomdin, Yoram Elihai, Miriam Briskin
Life of this patent
Estimating the term…
Abstract
Processes for pictorial representation by data compression including dividing a picture into regions of designated brightness values, fixing a characteristic scale for each region representing a number of pixels, dividing each region into cells, identifying each cell representing the basic structure of each cell by way of models, and storing and/or transmitting the data for each such model in the form of primary compression of the picture.
Claims28
A process of picture representation by data compression which comprises the steps of:
- subdividing the picture into regions;
- registering for each region a set of brightness values;
- fixing for each region a characteristic scale in terms of a number of pixels;
- dividing each region into cells, each of said cells comprising a number of pixels defined by two coordinates, said cells having a linear dimension in the order of said characteristic scale;
- identifying in each cell basic structures chosen from among smooth areas, positive and negative hills, and curvilinear structures chosen from among edges and ridges;
- constructing for said curvilinear structures geometric models comprising lines approximating the center lines of said structures and parameters defining the profiles of said structures;
- associating to each of said smooth areas, positive and negative hills and geometric models of curvilinear structures a mathematical model;
- condensing said mathematical models to define a global mathematical model for the cell;
- quantizing and encoding the data defining said global mathematical model;
- and
- storing and/or transmitting said data as representing the primary compression for the picture.
A process of representing a picture by data, comprising the steps of:
- subdividing the picture into regions;
- registering for each region a set of brightness values, said set of brightness values collectively defining a brightness function;
- dividing each region into cells, each of said cells comprising a number of pixels having a location defined by two coordinates;
- identifying in each said cell basic structures chosen among:
- (a) background areas wherein the values of said brightness function change slowly;
- (b) curvilinear edges having a center line wherein on one side of said edge the values of said brightness function undergo a sharp change;
- (c) curvilinear ridges having a center line, wherein the cross-sectional profile perpendicular to said center line is a bell-shaped curve;
- and
- (d) hills wherein the brightness function value is a maximum or minimum and decreases or increases, respectively, in all directions from said hills;
- constructing geometric models for those basic structures which are curvilinear, said geometric models comprising lines approximating the center lines of said structures and parameters defining the profiles of said structures;
- associating a mathematical model with each of said background areas, hills and geometric models;
- condensing said mathematical models to define a global mathematical model for the cell;
- quantizing and encoding the data defining said global mathematical model;
- and
- storing and/or transmitting said data.
Dependencies are inferred from the claim text.