二维卡尔曼滤波的多源信息序贯式融合去噪方法

图像去噪方法主要是基于单传感器进行研究的,单传感器的信息已不能满足图像处理的需求。因此文章首先建立了多传感器下的二维线性离散系统的状态空间模型,然后对图像进行2DKF去噪,最后采用多传感器的序贯式融合方法取得结果来满足图像处理的需求。

The main method of image denoising is addressed on the basis the single sensor image information. However, the sin-gle algorithm of a single sensor cannot meet the high accuracy need of image recognition tasks. In this paper, a two dimensional lin-ear discrete state space model of multi-sensor image system is established, firstly. And then the 2D Kalman filtering algorithm is uti-lized to complete the image denoising for each sensor. Finally, the multi-sensor information fusion in sequential fusion is utilized tofusion the results of each sensor to meet the needs of image processing.