The Research Of Navigate For Tea-plucking Machine Based On Machine Vision

China is the birthplace of tea,and it is the largest tea cultivation,tea production and tea export country in the world.At present tea picking in our country relies on the manual hand picking,or using the portable and knapsack tea-plucking machine mainly.This way of harvest needs to consume huge manpower,and lead to incomplete cutting,great losses to the tea tree.For the cutter and tea ridge is easy to appear the deviation problem when driving the teaplucking machine,this article designs a navigation method for tea-plucking machine based on machine vision.Firstly,this thesis summarized the research present situation in the field of agricultural based on machine vision.Researching and inducting the visual navigation for agricultural machinery research by domestic and foreign scholars.On this basis,according to the characteristics of the rea garden design the corresponding method to calculate the navigation line based on machine vision.This thesis is mainly focused on the following aspects:(1)Firstly,the road structure information is used to calibrate the external parameter of the camera.Determine the relative position between the camera and the tea-plucking machine.Secondly,combine the internal and external calibration parameters results to make coordinate mapping relationship.This relationship is between the camera imaging plane and the threedimensional coordinate system which is the center of the tea-plucking machine.Then,according to the height and width of the tea prior information,product the grid by a certain rule on the interested area.And project to the 2-D image plane based on the coordinate mapping.Form the new rectangular window which is small in the fore,and big in the distance.It provides the basis for image processing follow-up.(2)For the tea-plucking machine navigation method based on machine vision is designed in this thesis,researching a series of image preprocessing algorithm matching to the characteristics of collected image,like image graying,the local OTSU thresholds method to segment image,and morphological filtering.(3)For the influence of the light shadow of the complex outdoor environment,compare with general shadow removal algorithm,and proposes a simple,rapid,and adaptive shadow removal algorithm.Process the first OTSU threshold segmentation on the interest areas,and remove the tea that is under the light and without the shadow.The remaining area includes tea that under the shadow and other low luminance value areas what is the tea ridge and the hollow between tea.The second OTSU threshold segmentation is processed in these areas.Further remove the tea under the shadow,and make sure only tea ridge and tea hollow are left over.(4)For the rest of the tea ridge and tea hollow,this article design the corresponding edge points extraction algorithm based on space constraints,and remove the interference of the tea hollow.An improved Hough line transformation algorithm act on the edge points.The algorithm can fit edge line faster and more accurately.In this thesis,a tea-plucking machine navigation system is designed and realized using Java language.The main modules and related methods in such a system are introduced in detail.The last part of the article is used to summarize the research and the experimental results.Experiments show that the proposed system can find the center line of the tea ridge in real-time and accurately to a certain extent.Tip the driver aimed at the center line not only reduce the driving load,but also improve the integrity of tea cutting,reduce the damage of tea tree.