基于时空信息的自适应数字伪装研究

Research on adaptive digital camouflage based on spatio-temporal information

针对自适应伪装的关键性技术进行研究,以可变视角及动态环境场景为前提,采用图像处理和机器学习的方法,实现伪装目标的自适应隐身。在数码迷彩设计思想的前提下设计了一种多步骤的自适应数字伪装技术,通过从每帧图像上提取环境信息,利用颜色丰富度指标和k均值聚类算法提取主要颜色信息,使用基于贪婪流动场的迭代算法获取随机数码纹理,最后将颜色信息和纹理信息融合作为该帧图像的伪装。实验表明该算法对于伪装目标静止和运动过程有良好的伪装效果。这项研究对于自然探索、娱乐生活、潜伏侦查、国家安防和科学研究具有重大意义。

The key technologies of adaptive camouflage were studied.On the premise of variable viewing angle and dynamic environment scene,the adaptive camouflage of camouflage targets was realized by using image processing and machine learning methods.Based on the idea of Digital Camouflage Design,a multi-step adaptive digital camouflage technology was designed.By extracting environmental information from each frame of image,the color richness index and k-means clustering algorithm were used to extract the main color information,and the iterative algorithm based on greedy flow field was used to obtain the random digital texture.The color information and texture information were fused as the camouflage of the frame image.Experiments show that the algorithm has a good camouflage effect for the static and moving process of camouflaged targets.This research is of great significance for natural exploration,entertainment,latent investigation,national security and scientific research.