A New Image Fusion Algorithm for Recognition Capability Enhancements

zhang yongmei,ma li,liu wenkai
DOI: https://doi.org/10.1007/978-94-007-2169-2_29
2012-01-01
Abstract:Image fusion principles have been widely used in application of imaging remote sensing as an effective means of synergistic information combination. Multi-sensor data fusion can obtain much more and more exact information than single sensor. A sensor often with difficulty realizes target identification, multi-sensors can quite easily identify the goal. A new image fusion algorithm between multi-spectral and panchromatic images based on object recognition is presented, which focuses on the fusion rules of wavelet coefficients and the coherent selection. The algorithm makes full use of different characters of multi-spectral and panchromatic images, which can enhance recognition capability. Experiment data indicate this algorithm improves the fusion quality, and it can benefit the correct target classification. Therefore, the approach proposed is an effective method for target recognition.
What problem does this paper attempt to address?