Research On The Vision-aided Navigation Technology Of Automatic Weed Cleaning Workboat

China is a big consumer of crab and other aquatic products with a huge farming area and consumption market.However,compared with developed countries,China’s aquaculture industry exists problems of backward production mode,high production cost,small production scale and low level of industrialization.As the cleaning and maintenance of the aquatic plant of the crab breeding are all done by the manual work,it is necessary to develop a fully automatic weed cleaning workboat to reduce the labor intensity and the cost of breeding.In this paper,a vision-aided navigation system based on GPS positioning is designed to realize the automatic navigation function of fully automatic operation ship.The main work of this paper is as follows:1.Using the paddlewheel powered platform as the research object,a vision-aided navigation system based on GPS positioning is constructed.The design idea and hardware platform of vision navigation system are introduced in this paper.2.The image of water surface is processed in digital image processing technology.The first step is the pre-processing of the image,including noise filtering and grayscale processing,which lays the foundation for the subsequent image segmentation based on threshold.Then,the segmentation of the target(aquatic weed)and background(water surface)is completed,and the effective navigation area is obtained by the methods of mathematical morphology and the threshold area reservation.3.A navigation reference line is extracted from the effective navigation area.Firstly,the classical methods of least squares fitting and Hough transform line detection are used in the experiment.Besides,an improved linear fast detection algorithm is proposed to extract the baseline,which served as the workboat navigation path based on machine vision.4.Zhang’s calibration method is used to calibrate the CCD camera and obtain its intrinsic and extrinsic parameters in this paper.Based on the camera imaging model and the internal and external parameters of the camera,the navigation reference line in the image coordinate system is converted into the world coordinate system and the course angle can be calculated.5.The Kalman filter is used to fuse the navigation parameters of the vision navigation module and the GPS positioning module to improve the positioning accuracy of the automatic navigation system.Furthermore,a simulation experiment is designed to validate the method.The results show that the maximum deviation,average deviation and standard error of the experimental data are respectively decreased by 42%,65% and 67%.