Camouflage Evaluation And Its Application In Camouflage Texture Design

Digital camouflage as one of the most popular military camouflage technology plays a pivotal role in the fight against high-tech detection and attack weapon systems. How to further improve the performance of digital camouflage and design with a high-quality, anti-reconnaissance camouflage has important scientific significance and military application. The thesis focuses on digital camouflage performance evaluation and its feedback applications in the camouflage design.(1) Based on visual perception theory, the visual saliency is calculated to evaluate camouflage performance. In this thesis, a systemic and rational method for evaluation of camouflage texture designing is proposed, which fuses background-related features and background-unrelated features. In addition, we combine mosaic feature and saliency feature of camouflaged regions to improve the accuracy of camouflage performance evaluation. A set of subjective evaluation experiments are carried out as the baseline of our proposed evaluation model. Experimental results show that our method accords with the human visual perception mechanism(2) Evaluation of camouflage not only as an effective camouflage performance evaluation methods, and more importantly, as a feedback mechanism to regulate and optimize digital camouflage design process. Currently camouflage design is a forward tentative design. Up to now, we have not seen camouflage design based on "Design-Evaluation- Feedback- Design" closed-loop, and the quality of camouflage performance is unstable. Moreover, we use the performance evaluation as a constraint to optimize the design of camouflage textures, which improves the quality of camouflage design.