| 摘要 |
<p>An image fusion process performs respective pyramid decompositions (814, 816) of input images (810, 812). The components of the pyramid decomposition are processed (818, 820) for saliency to form respective saliency pyramids (822, 824). The salience pyramids are then combined (826) to form a masking pyramid (828) that defines which components of the two input image pyramids (814, 816) are to be combined to form the merged image. The masking pyramid is then used (830) to construct a merged image (832) from the pyramid decompositions of the input images.</p> |
