Accessibility navigation


Research on constructing difference-features to guide the fusion of dual-modal infrared images

Hu, P., Yang, F., Wei, H., Ji, L. and Wang, X. (2019) Research on constructing difference-features to guide the fusion of dual-modal infrared images. Infrared Physics & Technology, 102 (1). 102994. ISSN 1350-4495

Full text not archived in this repository.

It is advisable to refer to the publisher's version if you intend to cite from this work. See Guidance on citing.

To link to this item DOI: 10.1016/j.infrared.2019.102994

Abstract/Summary

The dual-modal infrared intensity and polarization images have great fusion val-ue due to their unique imaging advantages and rich complementary information between them. Using the complementary information of dual-modal infrared im-ages to guide fusion is the premise of configuring optimal fusion strategy (such as configuration of algorithms, selection of fusion rules, etc.). But the obvious drawback is that the description of the complementary information of dual-modal infrared images has always relied on the qualitative analysis of prior knowledge and experience summary. The complementary information identified by qualitative analysis is not only random and unpredictable, but also difficult to provide quantitative reference indexes for automated and intelligent fusion sys-tems. To solve this problem, this paper proposes to construct difference-features to quantitatively describe the complementary information of dual-modal infrared image and then three essential attributes that difference-features must possess are given from three perspectives: Difference in description, Independence in de-scription and Validity of guiding fusion. Furthermore, the imaging mechanism and image characteristics of dual-modal infrared images are analyzed and the sa-lient complementary information of the two types of images is identified as Brightness, Edge and Texture. Meanwhile, six difference-features are chosen or constructed to quantitatively describe these three types of complementary infor-mation. Namely, choosing average energy (AE) and information abundance (IA) to describe the complementarity of brightness difference; constructing edge strength (ES) and edge abundance (EA) to describe the complementarity of edge difference; choosing Tamura contrast (TC) and constructing visual index (VI) to describe the complementarity of texture difference. Experiments show that the six difference-features are independent of each other in describing the comple-mentary information of dual-modal infrared images. They cannot only effective-ly represent the complementary information of the difference between dual-modal infrared images, but also directly reflect the fusion performance by using them to guide fusion. Therefore, it has practical feasibility and important guiding significance to use them to guide fusion.

Item Type:Article
Refereed:Yes
Divisions:Faculty of Science > School of Mathematical, Physical and Computational Sciences > Department of Computer Science
ID Code:91414
Publisher:Elsevier

University Staff: Request a correction | Centaur Editors: Update this record

Page navigation