基于分水岭方法的数码迷彩设计

摘要

迷彩是对抗现代侦查与精确打击,有效提高军事目标战场生存能力的防护手段之一,是作战保障的重要内容.数码迷彩能够有效破坏伪装目标的外形,使不同色彩间的边界模糊、破碎,具有良好的伪装隐身效果.文章提出了一种新的生成数码迷彩算法,它先对图案使用分水岭方法分割,确定出大致图案,然后提取出背景主色和确定数码迷彩斑块的大小,最后生成数码迷彩.由于图像本身会存在噪声,在分割时会造成许多虚假边缘,即过分分割现象.为了避免产生过分分割现象,文中采用预处理滤波的方法,在图像应用分水岭算法前滤除图像噪声.实验表明,利用此算法生成的数码迷彩相对于传统迷彩能够更好地与自然背景相融合,达到良好的光学伪装效果.%Camouflage as one of the effective way that against modern detection and precision strike and improving the survival ability of military targets is important content of combat. The digital camouflage can break the boundary of a camouflaged target between different col-ors, and make it blurred and broken, so as to attain a better camouflaged effect This paper proposes a new algorithm of camouflage design-it would first segment image for determining the general pattern of camouflage, then extract color of background and determine the size of the digital camouflage plaques, finally create digital camouflage. The experimental result shows that, the digital camouflage is easier to blend with the natural background than the traditional camouflage, and achieves a good effect of the optical camouflage.