Gradient field divergence-based small target detection in infrared images

Infrared small target detection in complex cloud backgrounds has long been a research challenge. A novel robust method based on the divergence of gradient field is proposed to enhance and suppress background synchronously. The negative intensity (NIG field) matches with characteristics positive source. cluster, other hand, lacks this feature. First, NIG calculated target’s property from original image. values are then produce defined map (D map), which highlights regions while suppressing clutter regions. Meanwhile, local vectors angle measure (LVAM) operator designed distribution 8-neighbour eliminate false areas. Then, (LA map) obtained by measuring value for each patch field. In addition, divergence-local (D-LA as Hadamard product D LA map. Finally, we can easily obtain via constant alarm ratio D-LA performance evaluation results real image sequences show that satisfactory suppression detection. Moreover, comparative experiments outperforms conventional methods terms accuracy rate.