Multimedia Tools and Applications ( IF 3.6 ) Pub Date : 2020-05-16 , DOI: 10.1007/s11042-020-09002-5 Qi Jia , Wei Dong Xu , Jiang Hua Hu , Jun Liu , Xin Yang , Li Yan Zhu
Traditional camouflage pattern painting is not effective enough due to development of the high-tech reconnaissance. The recently proposed digital pattern painting, a new form of camouflage pattern painting, which has better performance in countering reconnaissance. In this study, we present a novel design and evaluation of digital camouflage pattern based on spot combination. Firstly, we compared three clustering algorithms to extract dominant colors. At the meantime, on the basis of spot database, a spot combination algorithm is proposed to generate camouflage spot pattern. Finally, we evaluate its camouflage effect using saliency map based on human visual perception mechanism, as well as detection time. The herein proposed digital camouflage pattern can effectively blend with surrounding background, significantly prolong the detection time, and thus has a good camouflage effect.
中文翻译:
通过点组合设计和评估数字迷彩图案
由于高科技侦察技术的发展,传统的迷彩图案绘画不够有效。最近提出的数字图案画,一种新形式的迷彩图案画,在抵抗侦察方面具有更好的性能。在这项研究中,我们提出了一种基于斑点组合的新颖的数字迷彩图案设计和评估。首先,我们比较了三种聚类算法以提取主色。同时,在斑点数据库的基础上,提出了一种斑点组合算法来产生伪装斑点图案。最后,我们使用基于人类视觉感知机制的显着性图以及检测时间来评估其伪装效果。本文提出的数字迷彩图案可以有效融合周围的背景,大大延长检测时间,