论文部分内容阅读
为提高农机在果园复杂环境下作业精度,降低驾驶员劳动强度,设计了精准辅助导航系统。通过北斗差分定位设备获取农业机械的经纬度、航向角度以及行驶速度等数据,根据农田信息规划出全局行驶路径,实时计算农业机械行驶过程中的航向偏角、行驶距离和耕作面积并绘制行驶轨迹,根据观察航向偏角和行驶轨迹,驾驶员来修正作业机械使农机按规划路径行驶。系统采用基于红外散斑的深度立体视觉相机,降低自然光线限制,以此获取农业机械前方深度图像,根据视觉算法对图像中的作物或障碍加以检测,并计算其到农业机械和路径中心线距离,判断是否需要避障并提示驾驶员应调整方向,当调整方向进行避障后返回到已规划路径继续行驶。试验结果表明,辅助导航系统在直线行驶试验时横向行驶最大偏差小于5 cm,能够辅助驾驶员按规划路径精准行驶;基于立体视觉的障碍检测测距误差范围在4%以内,且处理每帧图像耗时最大为40 ms,实时性很高,能高效辅助驾驶员进行精确避障。
In order to improve the precision of farm machinery in orchard under complex environment and reduce the labor intensity of the driver, a precise auxiliary navigation system is designed. Obtain the latitude, longitude, heading angle and driving speed data of agricultural machinery through Beidou differential positioning equipment. According to farmland information, we can plan out the global driving route and calculate the heading angle, driving distance and cultivated area in real time, According to the observed course angle and driving trajectory, the driver to amend the operating machinery to farm machinery according to the planned route. The system uses a depth stereo vision camera based on infrared speckle to reduce the limitation of natural light to obtain the depth image of the front of the agricultural machinery and detect the crop or obstacle in the image according to the visual algorithm and calculate the distance from the center line of the agricultural machinery and the path , To determine whether the obstacle avoidance is required and prompts the driver to adjust the direction, and when the direction is adjusted to avoid obstacle, return to the planned route to continue driving. The experimental results show that the maximum deviation of the horizontal navigation during the driving test is less than 5 cm, which can help the driver to drive accurately according to the planned route. The distance error based on stereo vision is within 4%, and the image of each frame is processed Time-consuming up to 40 ms, high real-time, can effectively assist the driver to accurately avoid obstacles.