Infrared small target detection based on joint local contrast measures

Fast and accurate detection of dim and small targets is a key feature in infrared (IR) search and tracking systems. Small targets with no obvious features are usually submerged in complex backgrounds and clutter, causing low detection rates for most methods. In this paper, a novel joint constraint local contrast measure algorithm is proposed to detect small IR targets. It consists of two modules. First, we define a ratio-difference measure to enhance the small target and suppress the background. Second, a constrained difference measure is defined to suppress clutter and enhance the target. The two contrast measures are combined to obtain the saliency map. Finally, an adaptive threshold is calculated to extract the target. Experiments on a series of real IR images and sequences demonstrate that the proposed method can achieve better detection performance than other state-of-the-art methods.