一种基于微分不变量的图像变化检测方法

配准误差、噪声干扰和照度变化是影响变化检测性能的主要因素,利用图像结构信息进行变化检测,可以有效地克服这些因素的影响.文中提出了一种利用微分不变量描述图像结构信息并进行变化检测的方法.微分不变量具有平移和旋转不变性,并且对噪声具有较强的鲁棒性.首先利用微分不变量构造特征描述子,然后在一个搜索窗内计算各描述子之间的Mahalanobis距离,取其最小值并与阈值相比作变化检测.实验证明,所提出的算法对噪声干扰和配准误差都有较强的鲁棒性.

An Image Change Detection Algorithm Based on Differential Invariants

Misregistration error, noise and illumination variance are the main factors affecting the performance of the change detection algorithm. The use of image structure information in change detection can make the algorithm robust to these. This paper proposed a new method to describe the image structure information and make change detection by using the differential invariants, which are robust to the noise and invariant to the translation and rotation. The differential invariants are firstly used to form local invariant feature, then the invariant features from two images in a search window are compared by calculating the Mahalanobis distance between two features, and finally a threshold of the distance is used to decide whether or not changes take place. Experimental results show that the algorithm proposed in this paper is robust to noise and misregistration error.