一种改进的代表点匹配算法在稳像技术中的应用

提出一种用于电子稳像技术改进的代表点匹配算法。通过合理的划分区域,保证在区域内各点运动的一致性,利用代表点匹配算法确定各区域内的运动矢量,并把其代入变换模型可得关于旋转、平移和变焦等信息的线性方程组,求其方程的解得全局运动矢量。利用该方法的特点是,具有速度快,并能求出平移运动和旋转运动参数,实现图像运动的平移和旋转补偿。实验表明,改进的稳像算法对于提高动态图像的稳定性具有较好的效果。

Application of improved representative point match algorithm in electronic image stabilization

An improved representative point match algorithm applied to electric digital image stabilization was presented. Through dividing image into many blocks in reason, every pixel motion in each region can make sure uniform. Then local motion vector of block was determined by representative point match algorithm and was applied to given motion model to obtained globe motion vector between the reference frame and the current frames by solution of a set of linear equations. The method not only has the characteristic of high processing image sequences but also can detect the translation and rotation of the image and realize motion compensation. The experiments show that improved algorithm has a better effect for improving dynamic image stabilization.