摘要
飞机尾旋试验是研究飞机进入尾旋特征以及提出如何改出尾旋的方法。传统测量飞机尾旋运动姿态的实验自动化程度低、人工操作复杂费时,因此针对这些问题开发了一套基于特征点测量的改进方法。主要是基于机器视觉的三维测量原理,能动态测量飞机模型的三维特征点并计算出模型的各个姿态角度。在标记检测时采用了新的标记检测方法,提高了标记布局及图像识别的灵活度;姿态角的计算是根据坐标系转换而得到的。根据改进方法研制的系统操作简单、整个测量过程只需30min左右;并且数据有效率控制在90%以上;姿态角的精度控制在±1°以内。该系统在中国空气动力研究与发展中心立式风洞中的飞机自由尾旋实验中得到了很好的应用。
The airplane free-spin test is to study spin characteristics and the spin recovery method. The traditional measurement method of the spin model attitude is nonautomatic, com- plex and time-consuming for manual operation. To solve these problems, an improved system based on feature points is completed. This system can dynamically measure the three-dimensional coordinates of feature points on the aircraft model, then the attitude angles of airplane model are calculated by the theory of computer vision. A new method of feature point detection is proposed, which can raise the flexibility of feature points' layout and the accuracy of image recognition. The attitude angles are calculated through coordinate conversion. For this new system the rate of data efficient are above 90 %, and the accuracy of posture angle are less than 4-1 degree. Besides, this measuring process only takes 30 minutes and can be easily operated.
出处
《实验流体力学》
CAS
CSCD
北大核心
2013年第2期77-81,共5页
Journal of Experiments in Fluid Mechanics
基金
四川省科技厅国际合作资助项目(项目编号:2009HH0023)
关键词
飞机尾旋
机器视觉
三维重建
标记检测
free-spin
machine vision
3-D reconstruction
mark detection