论文部分内容阅读
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。
An improved representative point matching algorithm for electronic image stabilization is proposed. By reasonably dividing the regions, the consistency of motion at each point in the region can be ensured. The representative point matching algorithm can be used to determine the motion vectors in each region. The linear equations of rotation, translation and zooming can be obtained by substituting them into the transformation model , Find the solution to the global motion vector equation. The characteristic of this method is that it has the advantages of fast speed, and can obtain the translational and rotational motion parameters to realize the translation and rotation compensation of image motion. Experiments show that the improved image stabilization algorithm has good effect on improving the stability of dynamic images.