靶场光测图像实时判读系统设计与方法

System design and method research for optical measurement images real-time interpretation in test ranges

研究了靶场光学测量图像实时判读的若干关键技术,并据此设计实现了一套实时判读解算系统。目标自动捕获、跟踪和判读特征高精度定位是实时判读的难点问题,分别采用了形状相似度分析、在线区域跟踪学习、迭代轮廓匹配定位多判读点等技术;在此基础上,根据实时处理的时效性要求高、可并行程度强以及流程时序严格等特点,采纳分布式架构融合业务流程管理的软件体系结构设计模式,实现了数据与流程的集中管理、多站图像并行判读、数据驱动的解算方法自动选择与友好的用户界面。以航空靶场空中发射试验任务为例,通过实验验证了系统的可行性和时效性。该系统可扩展应用于各类靶场试验,满足用户实时获取并分析武器试验数据的需求。

Some key issues of real-time interpretation for optical measurement images in test ranges to design were researched and a real-time interpretation and estimation system was realized. Automatic Target capture,tracking and high precision location for interpretation features are the questions under probe. Shape similarity analysis,online region tracking and learning,and iterative contour registration were respectively used to locate multiple interpretation points. Based on the characteristics of real-time processing,such as high efficiency requirement,strong parallelism capability, and rigorous sequence in time,the software architecture design pattern of distributed framework with business process management was adopted to achieve the centralized administration of data and process,multiple sites parallel interpretation,data-driven auto estimation method selection and friendly user interface. Take the aerial launching tests of the air ranges as an example,the feasibility and efficiency of the system was examined by the experiment. The system can be extended to various test ranges to meet the users＇ need of weapon test data＇s real-time gain and analysis.