以尺度类似于时间,对具有不同分辨率的多幅图像建立起状态方程和观测方程;以标准Kalman滤波为工具,将具有不同性能与特点的图像进行融合,并给出了分块快速算法。利用估计误差绝对值均值对融合的性能进行了评估.多组实验与分析表明:所提出的图像融合算法不仅能有效的去除噪声和提高图像分辨率,而且通过图像融合,能够大大改善存在部分遮挡和恶劣天气等影响下获取的存在灰度、对比度变化的图像的性能。
By treating scales as time, the system equations are established, which combines multiple images with different resolutions. Based on Kalman filtering, images with different characteristics and performances are fused effectively. By dividing images into blocks, a fast algorithm with low computation complex is presented. The performance of the image fusion algorithm is evaluated by the absolute values of the estimation errors Theoretical analysis and simulation experiments demonstrate that the proposed method is effective in the fusion of both Gaussian noisy images and images with partial occlusion, intensity or contrast changes.