Browse Prior Art Database

Point-and-click method of selecting color objects with region growing process for color correction

IP.com Disclosure Number: IPCOM000199427D
Publication Date: 2010-Sep-02
Document File: 7 page(s) / 273K

Publishing Venue

The IP.com Prior Art Database

Abstract

This idea proposes a method of selecting color objects with a region growing process. The region growing starts from seeds corresponding to the user selection (through point-and-click) and stops where a color cost function reaches a certain threshold. The color cost function includes factors such as the number of pixels in the neighborhood that belong to the object, the strength of color edge which is determined by the largest Eigen value of the Laplacian matrix, a covariance of the spatial gradient matrix, and the color difference between the seed pixel and the pixel of interest. Optionally the Laplacian edge map can be presented to the user to guide the object selection operation. The proposed leak-proof object selection method enables customized color correction on the object of interest without affecting other areas.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 33% of the total text.

Point-and-click method of selecting color objects with region growing process for color correction

This idea proposes a method of selecting color objects with a region growing process. The region growing starts from seeds corresponding to the user selection (through point-and-click) and stops where a color cost function reaches a certain threshold. The color cost function includes factors such as the number of pixels in the neighborhood that belong to the object, the strength of color edge which is determined by the largest Eigen value of the Laplacian matrix, a covariance of the spatial gradient matrix, and the color difference between the seed pixel and the pixel of interest. Optionally the Laplacian edge map can be presented to the user to guide the object selection operation. The proposed leak-proof object selection method enables customized color correction on the object of interest without affecting other areas.

In color correction, there is a need to correct colors for specific regions of the image since a global rendering Lookup Table (LUT) may not be optimized local to a selected color object (e.g. face, sky, etc.). This feature would be particularly valuable for correcting high quality digital images (photographic quality) on a selective basis using multi-dimensional color correction LUT’s specifically tuned for those color objects.

This idea proposes to use a region growing process with a spatial (i.e. 2D) color gradient matrix in RGB, L*a*b*, or CMYK color space. A grid with a Laplacian edge map is constructed for the whole image a priori using the spatial color gradient matrix. The Laplacian edge map (a template), if needed, is shown on the graphical user interface (GUI) to guide the user to point at his or her object of interest. After clicking the mouse, the algorithm (part of the proposal) will grow the region until it coincides with the Laplacian edge map. A color constrained cost function is introduced to prevent leakage through mild transitions between different objects. Once the object is selected, a customized rendering LUT is applied to improve the color quality of the object.

Background:

During the color tuning stage, particularly for special images in a job, customers are interested in tuning the images for improving their color rendition quality. There are no good object selection methods available in the literature that can enhance user’s experience and improve object selection and rendering experience. We propose a novel technique based on color constrained spatial gradient matrix with region growing cost function.

Proposal:

The point-and-click method described in this proposal performs object segmentation of the color object that is of interest to the user. The object segmentation method is designed for color images, which is novel when compared to object segmentation performed on gray scale images. As is well known in the color imaging community, different colors in three-color space may result in the same lum...