Research On Obstacle Recognition Method Based On Millimeter Wave Radar And Machine Vision

With the expansion of Unmanned Aerial Vehicle application,obstacle recognition is becoming the focus of people's research due to the complex flight environment.Since the limitation of single sensor,a model integrating Millimeter Wave radar and visual system is proposed: Radar is used to compensate for the limitation of depth measurement.The machine vision system can obtain the contour,size and shape of obstacles.Firstly,the hardware scheme was formulated to ensure the accuracy.Image preprocessing and processing model was established: the least square method was used to fit the curve,select parameters to improve the image quality and reduce the influence of environment and exposure.Through color value extraction,image binarization,morphological processing and other algorithms,ensure the efficiency,reduce the interference of other elements and improve the accuracy.Canny detection was used to get relatively smooth contour information.With the help of coordinate system,the area of obstacles and the distance between obstacles can be obtained.Secondly,in order to solve the problem that the information acquisition was out of sync,synchronize spatial and temporal information,the position and coordinate transformation of multiple sensors were combined.According to the sampling frequency of selected sensors and thread synchronization,the information was guaranteed in time.Finally,the simulation and verification were carried out in the plant protection.The hardware was selected,the actual distance,azimuth relation,outline shape of obstacles,size of the obstacles and the horizontal distance between them were obtained,these information help us judge whether the plane can pass smoothly from the two obstacles.By controlling Unmanned Aerial Vehicle to a certain height,measuring the actual distance,controlling drone continues to fly forward,the accuracy of the results was proved,which provides a practical and reliable basis for the implementation of obstacle avoidance strategy.Figure 42;Table 5;Reference 59...