After completion of a planting protection operationthe UAV needs to land autonomously.The landing of conventional UAVs is usually assisted by their own GPS systembut big errors will occur when the system is applied to small UAVs in precise positioning.To address this issuefor the situation when the UAV needs to land at a fixed pointa fixed-point recognition method based on the machine vision system is studied.Firstlya specific landmark is set as the fixed pointand different color values in the image captured by the machine vision system are extracted.After color value extractionexcept the color of the fixed pointthe interference information of all the other colors is filtered.Thenthe shape information of all the elements in the image is analyzed.After the shape analysisall the elements whose shape doesnt accord with the shape of the fixed point are excluded.The contour of the landmark is identified through Canny edge detection.A plane coordinate system is set upand the specific location of the landmark in the coordinate system is calculated by using the pixel information obtained from edge detection.Finallyexperimental results show that the machine vision system can assist the UAV to precisely land at the fixed point.