| With the use of large-scale aseptic mushroom production workshop,mushroom production has been greatly increased.However,the mushroom picking operation relies on manual work for a long time,which has high labor cost,low efficiency and difficult quality management.Therefore,it is urgent to realize the automation of mushroom picking.On the basis of summarizing the advantages and disadvantages of agricultural picking robots at home and abroad,in view of large-scale aseptic mushroom production,taking brown mushroom planted in seedbed as the research object,the key technology of automatic mushroom picking based on machine vision was studied,and on this basis,an automatic mushroom picking system based on machine vision was developed.The main research contents are as follows.:(1)Analysis of the problem of automatic mushroom picking in large-scale aseptic mushroom production workshop,the architecture of automatic mushroom picking system based on machine vision and the process of automatic mushroom picking were designed.The mobile platform and picking manipulator for automatic mushroom picking were designed based on the existing truss structure of mushroom seedbed.(2)Aiming at the problem of complex background filtering of mushroom image,a complex background filtering algorithm based on Harris corner is designed.After denoising and enhancing the image,the effective filtering of complex background of mushroom image is realized by using the difference between the foreground of mushroom image and the density of Harris corner of background.(3)To solve the problem of sticky mushroom separation under different scales,a marker point search algorithm based on iteration method was proposed,and then marker watershed algorithm was used to accurately separate sticky mushrooms with different scales.(4)Aiming at the problem of calculating the picking angle and planning the picking path of the manipulator,this paper analyzed the position relationship between the sticky mushrooms,and then constructed the distribution model of the sticky mushrooms.On this basis,the method of calculating the picking angle of the sticky mushrooms with the constraint of not damaging the neighborhood was studied.Finally,the picking path planning of all the mushrooms was carried out by the way of going back and forth.The genetic algorithm was used to optimize the picking path.(5)On the basis of the above research,an automatic mushroom picking system based on machine vision was built,and the image positioning accuracy and the overall performance of the system were tested.The test results show that the positioning accuracy of mushroom image can reach 1.5mm,the correct recognition rate of image is 89.63%,and the picking time of single mushroom is 12.8s,which can meet the requirements of enterprises for automatic mushroom picking. |
