Video Shake Detection

Video surveillance systems play an irreplaceable role in power scene monitoring. However, in the process of image acquisition, video surveillance equipment will inevitably have various external interference factors. This paper proposes an effective method for video shake detection. This method incorporates the classical BOW (bag-of-words) model into the traditional projection method to get the acquired feature representation in a higher level and make it more robust. In addition, we focus on the local characteristics of the image and use a block operation. The experimental results show the effectiveness of our method.