一种稳定船载电视侦察图像序列的灰度投影拟合算法

提出了一种新的船载电子稳像算法———灰度投影拟合算法,此算法包括:合理划分图像区域,保证区域内各点运动一致;利用灰度投影法确定每个区域内的图像平移量;利用拟合公式确定图像全局运动矢量。既保留了灰度投影算法的快速性,又保证了图像旋转量的准确确定,从而实现了图像灰度细节丰富但无明显特征目标的图像序列的稳定。对船载摄像系统获取的视频图像序列进行稳像处理的结果表明,该稳像算法稳像精度高,实时处理能力强,实际应用效果好。

Gray projection-fit algorithm for stabilizing image sequences acquired from ship-borne TV reconnaissance system

A new electronic image-stabilizing algorithm-gray projection-fit algorithm is introduced. This algorithms is composed of three parts:Firstly, the image is divided to many regions ,and make sure that every pixel motion in each region is uniform;Secondly, the translations of each region are obtained by the gray projection algorithm;Finally the global motion value of image is calculated by the fit formula according to the translations of each region. This algorithm not only overcomes the disadvantage of the gray projection algorithm that couldn't detect the rotation of the image, but also remains its high speediness to stabilize the image sequences with faint features and too gray details. The algorithm is shown how to process the image sequences taken on the ship and make it stabilized. The experimental result shows that the adoption of the image-stabilizing algorithm can really produce high image stabilizing precision, excellent processing capacity on real time and achieve good effects in actual application.