论文部分内容阅读
为了弥补金字塔变换算法分解时数据冗余较大、融合结果不理想的缺点,提出基于金字塔变换算法优化的遥感图像融合新算法。该算法运用金字塔分解构建金字塔序列,并根据先验知识赋予相应的权重系数,通过反复迭代重建遥感图像,再利用班德文克隆选择算法优化选择,在迭代可承受的范围内,自适应地修改选择权重系数,寻求和估计合适的融合参数来优化融合效果,从而避免金字塔变换算法的经验选择。为了突出本文算法的优点,实验运用金字塔变换法、遗传算法优化金字塔变换法和粒子群算法优化金字塔变换法进行比较,从视觉效果和数理统计两个方面分析评价融合质量。实验结果表明,本文算法更符合人类的视觉感知,有利于图像的分析和信息的提取。
In order to make up for the shortcomings of pyramid transformation algorithm decomposition, such as large data redundancy and unsatisfactory fusion result, a new remote sensing image fusion algorithm based on pyramid transformation algorithm is proposed. The algorithm uses pyramid decomposition to construct the pyramid sequence, and assigns the corresponding weight coefficients according to the priori knowledge. By repeatedly iteratively reconstructing the remote sensing images, the algorithm can be optimized and selected by using the Bandeaven clonal selection algorithm, and the algorithm can be adaptively modified within an iteratively acceptable range Select the weight coefficient, seek and estimate the appropriate fusion parameters to optimize the fusion effect, avoiding the empirical choice of the pyramid transform algorithm. In order to highlight the advantages of our algorithm, the experiment uses the pyramid transform method, the genetic algorithm optimized pyramid transform method and the particle swarm optimization algorithm pyramid transform method to compare, and evaluates the fusion quality from two aspects of visual effect and mathematical statistics. Experimental results show that this algorithm is more in line with human visual perception, which is good for image analysis and information extraction.