Process to extract regions of homogeneous color in a digital picture

• US 7,016,531 B1
• Filed: 02/01/2000
• Issued: 03/21/2006
• Est. Priority Date: 02/01/1999
• Status: Expired due to Fees

• Current Assignee: Gvbb Holdings SARL (Vislink Technologies, Inc.)
• Original Assignee: Thomson Licensing (Vantiva SA)
• Inventors: Tabatabai, Ali, Murching, Anil M., Naveen, Thumpudi
• Primary Examiner(s): Bali, Vikkram
• Assistant Examiner(s): LAROSE, COLIN M
• Application Number:
  US09/496,068
• Time in Patent Office: 2,240 Days
• Field of Search: 382/164, 382/165, 382/173
• US Class Current: 382/164
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 2207/20021 Dividing image into blocks, subimages or windows
• CPC Class Codes: G06T 2207/20152 Watershed segmentation
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/155 involving morphological operators
• CPC Class Codes: G06V 10/267 by performing operations on regions, e.g. growing, shrinking or watersheds

Abstract

A method of extracting regions of homogeneous color from a digital picture divides the digital picture into blocks and generates a feature vector for each block as a set of moments of the data for the block. The distance between the feature vector of each block and the feature vectors of the nearest neighboring blocks are determined using either a weighted Euclidean distance metric or a probability mass function-based distance metric. The maximum distance is the gradient value for the block, and the set of gradient values over all the blocks form a color gradient field. The gradient field is digitized and smoothed, and then segmented into regions of similar color characteristics using a watershed algorithm.

Claims

1. A method of extracting regions of homogenous color in a digital picture comprising the steps of:

dividing the digital picture into blocks wherein each block comprises a plurality of pixels; and

merging together spatially adjacent blocks that have similar color properties to extract the regions of homogenous color, wherein the merging step comprises the additional steps of;

extracting a feature vector for each block;

estimate a scalar gradient value for each block as a function of the feature vector, the set of gradient values defining a color gradient field;

digitizing the color gradient field;

preprocessing the digitized color gradient field to produce a smoothed color gradient field; and

segmenting the smoothed color gradient field with a watershed algorithm that divides the smoothed color gradient field into a set of spatially connected regions of homogenous color;

wherein

wherein the estimating step comprises the steps of;

obtaining distances between the feature vector of each block and the feature vectors of each neighboring block; and

selecting the maximum of the distance as the gradient value for the block; and

wherein the obtaining step comprises the steps of;

converting the feature vector of each block into a probability mass function-based representation for each color component;

computing distances between the probability mass function-based representations of each block and the corresponding probability mass function-based representations of each neighboring block; and

selecting the maximum distance of the probability mass function based representations as the gradient value for the block.