Image fusion using bi-directional similarity

Chunshan Bai; Xiaoyan Luo

doi:10.1117/12.2178932

8 May 2015 Image fusion using bi-directional similarity

Chunshan Bai, Xiaoyan Luo

Proceedings Volume 9508, Holography: Advances and Modern Trends IV; 95080Q (2015) https://doi.org/10.1117/12.2178932
Event: SPIE Optics + Optoelectronics, 2015, Prague, Czech Republic

Abstract

Infrared images are widely used in the practical applications to capture abundant information. However, it is still challenging to enhance the infrared image by the visual image. In this paper, we propose an effective method using bidirectional similarity. In the proposed method, we aim to find an optimal solution from many feasible solutions without introducing intermediate image. We employ some priori constraints to meet the requirements of image fusion which can be detailed to preserve both good characteristics in the infrared image and spatial information in the visual image. In the iterative step, we use the matrix with the square of the difference between images to integrate the image holding most information. We call this matrix the bidirectional similarity distance. By the bidirectional similarity distance, we can get the transitive images. Then, we fuse the images according to the weight. Experimental results show that, compared to the traditional image fusion algorithm, fusion images from bidirectional similarity fusion algorithm have greatly improved in the subjective vision, entropy, structural similarity index measurement. We believe that the proposed scheme can have a wide applications.

Citation Download Citation

Chunshan Bai and Xiaoyan Luo "Image fusion using bi-directional similarity", Proc. SPIE 9508, Holography: Advances and Modern Trends IV, 95080Q (8 May 2015); https://doi.org/10.1117/12.2178932

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