Infrared small target detection algorithm based on LBF level set and temporal-spatial information

The moving target detection is not only one of the key techniques in intelligent monitoring system and modern warfare but also an important field of image processing. Therefore, the key to ensuring the success of the follow-up tasks is to find the target as soon as possible. Due to the complex target background and the uncertainty of the target motion state, there are the missed detection and false alarm phenomena in the target detection. So, this paper proposes a new algorithm for infrared small target detection based on LBF level set and temporal-spatial information. The algorithm firstly uses the level set evolution of local binary fitting model to extract the zone of infrared small targets, subsequently use the temporal and spatial information of small target to eliminate the pseudo target, then the real infrared small target is detected. The Experimental results show that the proposed algorithm not only accelerates the convergence speed but also improves the detection rate of small infrared target.