1: \begin{abstract}	

2: 	Many state-of-the-art methods have been proposed for infrared small target detection.

3: 	They work well on the images with homogeneous backgrounds and high-contrast targets.

4: 	However, when facing highly heterogeneous backgrounds, they would not perform very well, mainly due to: 1) the existence of strong edges and other interfering components, 2) not utilizing the priors fully. 

5: 	Inspired by this, we propose a novel method to exploit both local and non-local priors simultaneously. 

6: 	Firstly, we employ a new infrared patch-tensor (IPT) model to represent the image and preserve its spatial correlations.    

7: 	Exploiting the target sparse prior and background non-local self-correlation prior, the target-background separation is modeled as a robust low-rank tensor recovery problem. 

8: 	Moreover, with the help of the structure tensor and reweighted idea, we design an entry-wise local-structure-adaptive and sparsity enhancing weight to replace the globally constant weighting parameter. 

9: 	The decomposition could be achieved via the element-wise reweighted higher-order robust principal component analysis with an additional convergence condition according to the practical situation of target detection.  

10: 	Extensive experiments demonstrate that our model outperforms the other state-of-the-arts, in particular for the images with very dim targets and heavy clutters. 

11: \end{abstract}

12: