Based on laser displacement sensor technology

a new obstacle avoidance method is proposed for plant protection unmanned aerial vehicles（UAV）. The method includes the following three parts: data block extraction

obstacle basic parameter calculation and obstacle pattern recognition. Based on the validity of distance value

data block corresponding to obstacle is extracted from data sequence. The average angle

average distance and width are used as the basic parameters to describe the obstacle. The design of pattern recognition classifier based on the width of obstacle

maximum clearance of data block

jump times and variance. Experiments show that the proposed method can effectively detect the position and distance of the obstacle in unknown environment and make a judgment on obstacle type.