采用多尺度掩模消除干扰的反向组合算法

反向组合算法是最有效的图像对齐算法之一,但该算法抗干扰能力差.当输入图像部分被遮挡时,图像对齐效果变差.针对该问题,我们提出一种采用多尺度掩模消除干扰的反向组合算法.该算法采用自适应的方法设置初始掩模,再通过迭代判断和逐层分块来细化初始掩模,使得掩模能准确地设置在干扰区域上.实验结果表明,该算法既保留了原反向组合算法的优点,又提高了算法抗干扰的能力,使得反向组合算法能在更复杂的环境下进行图像对齐.

An Efficient Inverse Compositional Algorithm Using Multi-Scale Mask to Eliminate Disturbance

Inverse Compositional Algorithm is one of the most efficient image alignment algorithms, but the ability of anti-jamming of the algorithm needs to be improved. The efficiency of the algorithm becomes worse when the image is occluded. We propose an efficient inverse compositional algorithm with multi-scale mask to eliminate disturbance. An adaptive method is used to get the scale of the original mask first, then the mask scale becomes gradually small to fix on the right place by estimating and blocking in the iterative process. The experimental results show that our algorithm not only keeps the superiority of the original inverse composition algorithm, but also enhances the ability of anti-jamming. And the improved inverse composition algorithm can still be efficient under complex environments.