针对固体火箭发动机飞行试验中喷管排出的羽流烟雾光学信号的准确测量,应用数字图像处理技术提取羽流烟雾的静态和动态特征进行羽烟识别.为了滤除外场复杂环境对烟雾测量的干扰,设计了具有等间距的黑白条纹背景板.在黑条纹区域采用自适应的灰度阈值提取烟雾特征,在白条纹区域根据RGB和HSV颜色空间综合的方法提取烟雾特征,结合帧间差分法提取视频序列中烟雾区域的扩散边界.基于烟雾静态和动态特征综合的算法编写了烟雾处理软件,通过处理大量烟雾视频数据进行验证试验,证明羽流烟雾的识别能够有效提高光学信号测量的准确度.%Focus on accurate measurement of optical signals from the plume in solid rocket motor flight test,digital image processing technology is applied to extract plume static and dynamic characteris-tics,therefore,the plume-smoke can be identified. In order to filter out the interference of complex envi-ronment on smoke measurement, black and white striped background plate with equal spacing was de-signed. In the black stripe region, adaptive gray threshold is used to extract the smoke feature and the method of RGB and HIS color-space information synthesis is used in white stripe region. The diffusion boundary of smoke region in video sequences is extracted by the method of frame difference. The smoke processing software is compiled based on the algorithm for combining static and dynamic characteristics of smoke. Through a large number of smoke video data validation test,it is proved that plume-smoke recog-nition can improve the accuracy of optical signal measurement.
展开▼
