Abstract: The restoration of composite-frame images degraded by relative motion between a camera and an object is systematically studied in this paper. First, the edge “staircase effect” is eliminated through the displacement vector (i.e. motion extent and direction between two sub-fields) estimation based on a block-matching algorithm; Second, the Wiener filtering is applied to remove the motion blur in each sub-field; Finally, through restoring the blurred images captured by relative linear motion with different speed and distance between a camera and a doorplate mounted in a vehicle, the method is validated.

Key words: Motion blur; Composite frame; Interlaced images; Displacement vector estimation; Image restoration

摘要： 对复合帧图像运动模糊的复原问题进行了较为系统的研究。利用块匹配算法估计场间的位移矢量，即模糊范围和模糊方向以消除边缘“阶梯效应”；采用维纳滤波方法对场内的运动模糊图像进行了复原；经对一个CCD 照相机以不同的相对运动速度在不同的距离对一个门牌号摄取的模糊图像进行复原，证明该文方法的有效性。

关键词: 天津大学电气与自动化工程学院，天津300072)