Fusion of IR and Visual Images,
Abstract
Image fusion is the process where several independently recorded images are combined into a single image. This report describes the fusion of thermal IR and visual imagery. For image fusion an optimality criterion is defined. If the input images are assumed to be dependent the optimal fusion process is image blending. For independent input images pyramidal solutions such as Burt's algorithm are shown to be optimal. Several image sequences are processed with the techniques described in the report. Results presented in the report indicate that, for the images used, image fusion results in images with a higher information content. Hardware requirements for the algorithms presented are given. Real-time implementation on current plug-in DSP boards seems possible. An initial study is performed into Automatic Target Recognition based upon multiple input images. The results look promising.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1997
- Accession Number
- ADA326139
Entities
People
- K. Schutte
Organizations
- Netherlands Organisation for Applied Scientific Research