The purpose of multispectral image fusion is to produce a synthetic all - in - focus image from several images each with different focusing settings. Nevertheless, if the resolution of the resource images is small, the fused images with the conventional fusion method will not be of any quality, which will prevent additional image analysis even of the fused image. This paper proposes a fusion of sources images with frequency partition discrete cosine transformation (FP - DCT) and normalization. The FP - DCT is used to obtain and fuse the majority of the image characteristics. Normalization is used to improve an image's dynamic range. Several pairs of multispectral images have confirmed the efficacy of the proposed scheme and compared it visually and quantitatively to obtainable methods. It is established that none of the methods demonstrated consistency in performance measurements of all. In most cases, however, the proposed method has demonstrated good quality routine. Moreover, by the proposed method, the visual superiority of the fused image is higher than other methods.
Volume 11 | 11-Special Issue
Pages: 546-555
DOI: 10.5373/JARDCS/V11SP11/20193065