摘要
针对电力系统巡线无人机自主飞行与自动避障的应用要求,以高压输电线、钢塔为特定对象,提出一种基于视觉信息的航线识别策略,以及一种改进的栅格路径选择方法。定义障碍物代价地图矩阵,矩阵元素值表示栅格图两栅格之间是否关联以及关联度,利用摄像头作为传感器件采集对象信息数据,通过滤波、平滑处理后的数据构建栅格代价矩阵并实施刷新,再此基础上,增加了阻碍系数检测与判定条件。最后,在ARM系列的STM32F407硬件平台上进行了验证,经过实际飞行测试,获得了满意的效果。
On patrol uav autonomous flight of the power system and application requirements of automatic obstacle avoidance,with high voltage transmission line,steel tower for a particular object,put forward a kind of route recognition strategy based on visual information,and an improved method of grid path selection. Define obstacle map matrix,matrix element value grid figure whether correlation and correlation degree between two grid,using camera as the sensor data acquisition object information,after filtering and smoothing data to construct grid cost matrix and the implementation of the refresh,again,on the basis of increased the block coefficient of detection and determination conditions. Finally,the hardware platform of the ARM series STM32F407 was validated,and the actual flight test was satisfactory.
出处
《技术与市场》
2017年第8期31-32,35,共3页
Technology and Market
关键词
巡线
避障
栅格法
关联度
代价矩阵
Patrol
Obstacle avoidance
Grid method
Correlation
Cost matrix
