A scale adaptive compressed object tracking algorithm

Abstract

In order to solve the problem of compressive tracking algorithm which cannot adapt to the scale change in object tracking, a scale adaptive compressed object tracking algorithm is proposed. Firstly, we extract the low-dimensional gray and texture features of the target and its surrounding region by sparse measurement matrix which reduces the complexity of compute. Then the tracking task is formulated as a binary classification via support vector machine (SVM) classifier with online update in the compressed domain. Meanwhile, we utilize the classifier to obtain the target's position in new frame. In addition, we build the hamming distance between hash values of current target and original one to match the template to achieve adaptive template size. Numerous experimental results show the proposed algorithm can keep tracking effectively when the tracked object is under the situation of scale variation and illumination change.