Human vision based color edge detection

Ari Kim; Hong-suk Kim; Seung-ok Park

doi:10.1117/12.877122

25 January 2011 Human vision based color edge detection

Ari Kim, Hong-suk Kim, Seung-ok Park

Proceedings Volume 7866, Color Imaging XVI: Displaying, Processing, Hardcopy, and Applications; 78660I (2011) https://doi.org/10.1117/12.877122
Event: IS&T/SPIE Electronic Imaging, 2011, San Francisco Airport, California, United States

Abstract

Edge detection can be of great importance to image processing in various digital imaging applications such as digital television and camera. Therefore, extracting more accurate edge properties are significantly demanded for achieving a better image understanding. In vector gradient edge detection, absolute difference of RGB values between a center pixel value, and its neighborhood values are usually used, although such a device-dependent color space does not account for human visual characteristics well. The goal of this study is to test a variety of color difference equations and propose the most effective model that can be used for the purpose of color edge detection. Three of synthetic images generated using perceptibility threshold of the human visual system were used for objectively evaluate to 5 color difference equations studied in this paper. A set of 6 complex color images was also used to testing the 5 color difference equations psychophysically. The equations include ΔRGB, ΔE^* _ab, ΔE_CMC, CIEDE2000 (ΔE₀₀) and CIECAM02-UCS delta E (ΔE_CAM-UCS). Consequently, there were not significant performance variations observed between those 5 color difference equations for the purpose of edge detection. However, ΔE₀₀ and ΔE_CAM-UCS showed slightly higher mean opinion score (MOS) in detected edge information.

Citation Download Citation

Ari Kim, Hong-suk Kim, and Seung-ok Park "Human vision based color edge detection", Proc. SPIE 7866, Color Imaging XVI: Displaying, Processing, Hardcopy, and Applications, 78660I (25 January 2011); https://doi.org/10.1117/12.877122

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available