Contrast and distribution based saliency detection in infrared images

Saliency-based approaches has been well studied and successfully used in object detection for visible images. However, few researches have been done for saliency detection in infrared images, which are characterized with low resolution, SNR and contrast, fuzzy edge and lack of color features. In this paper, a contrast and distribution based saliency detection approach is proposed for infrared images. First, we develop an enhanced multi-scale saliency feature by improving the quality and contrast of the image in frequency domain. Second, luminance-distribution and gradient feature are explored to highlight the object with great gradient and compact distribution. Finally, by integrating the above two features, the final saliency map for infrared image were obtained. Experimental results on real infrared images demonstrate the effectiveness of the proposed approach against the state-of-the-art algorithms.