Retrieval and matching of color patterns based on a predetermined vocabulary and grammar

• US 20030044062A1
• Filed: 07/03/2002
• Published: 03/06/2003
• Est. Priority Date: 01/25/1999
• Status: Active Grant

• Current Assignee: WSOU Investments, LLC (WSOU Holdings, LLC)
• Original Assignee: Lucent Technologies, Inc. (Nokia Corporation)
• Inventors: Mojsilovic, Aleksandra, Ganapathy, S. Kicha, Safranek, Robert James, Hu, Jianying, Kovacevic, Jelena
• Granted Patent: US 6,732,119 B2
• Time in Patent Office: Days
• US Class Current: 382/165
• CPC Class Codes: G06F 16/5838 using colour
• CPC Class Codes: G06F 16/5862 using texture
• CPC Class Codes: G06F 18/24765 Rule-based classification
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99936 Pattern matching access
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing
• CPC Class Codes: Y10S 707/99948 Application of database or data structure, e.g. distributed, multimedia, or image

Abstract

The invention provides a perceptually-based system for pattern retrieval and matching, suitable for use in a wide variety of information processing applications. An illustrative embodiment of the system uses a predetermined vocabulary comprising one or more dimensions to extract color and texture information from an information signal, e.g., an image, selected by a user. The system then generates a distance measure characterizing the relationship of the selected image to another image stored in a database, by applying a grammar, comprising a set of predetermined rules, to the color and texture information extracted from the selected image and corresponding color and texture information associated with the stored image. The vocabulary may include dimensions such as overall color, directionality and orientation, regularity and placement, color purity, and pattern complexity and heaviness. The rules in the grammar may include equal pattern, overall appearance, similar pattern, and dominant color and general impression, with each of the rules expressed as a logical combination of values generated for one or more of the dimensions. The distance measure may include separate color and texture metrics characterizing the similarity of the respective color and texture of the two images being compared. The invention is also applicable to other types of information signals, such as sequences of video frames.

Claims

1. A method of processing information signals, the method comprising the steps of:

extracting color information from an information signal in order to generate a color distribution of the information signal; and

generating an achromatic pattern map for the information signal using the color distribution generated from the information signal, wherein the achromatic pattern map includes a plurality of dominant colors of the color information each mapped into corresponding gray scale levels, and wherein the dominant colors are mapped into the gray levels based on the pixel areas associated with the dominant colors.

2. A method of processing information signals, the method comprising the steps of:

extracting color information from an information signal, using a separate color codebook for each of a plurality of luminance levels in the information signal, wherein at least a subset of the color codebooks include a different number of colors for each of the luminance levels; and

generating a color distribution representative of the extracted color information.

3. A method of processing information signals, the method comprising the steps of:

extracting at least one of color information and texture information from a first information signal; and

generating at least one of a color metric and a texture metric for the first information signal;

wherein the color metric characterizes the relationship of the color information of the first information signal to color information of a second information signal, and is generated at least in part as a function of differences in dominant color and corresponding pixel area for each of a number of dominant colors in the first information signal; and

wherein the texture metric characterizes the relationship of the texture information of the first information signal to texture information of the second information signal, and is generated as a weighted function of at least a first measure indicative of a difference in mean edge density between the on signal and a second measure indicative of a on of the first and second information signal.