Study On An Adjustable Grafting Seeding Cutting Device Based On Machine Vision

Due to the freedom and diversity of the growth of the plug seedlings,the stem thickness,height,the degree of cotyledon expansion,and the growth point of the stem are very irregular,which is difficult for the automatic operation of the grafting machine.In this paper,aiming at the automatic adjustment of cutting point and cutting angle of scion and rootstock during the automatic grafting of vegetable grafting machine,this paper puts forward the research on the cutting and adjusting device of grafting seed based on visual inspection,and uses the visual detection technology to drive the manipulator,the cutting and adjusting device Gestational adjustment of grafted seedlings.Main research work included following aspects.(1)The overall design of visual servo grafting adjustment system.Combined with the functional requirements of the cutting and cutting equipment,the overall structure of the visual servo grafting system is designed,which is mainly composed of the visual part and the control part.Using 1280 × 960 video stream acquisition device,based on opencv capture video stream and directly into the memory processing stack,the test can be collected in 3.9364 s 60 images,acquisition frequency of 15.2422 frames / sec.At the same time,it analyzes the main control device and the visual feedback feedback communication mode of the control subsystem.The communication speed of the I2 C bus can be customized to meet the communication requirements of Raspberry Pi and Arduino.(2)Study on visual inspection method of grafted.According to the function of grafting and the cutting and cutting system of grafted seedlings,the characteristics of scion and rootstock were extracted respectively.The processing rate of the multi-thread parallel preprocessing is improved from 2.299 to 2.873 times,the performance is improved by 56.50% ~ 65.19%,and the rate of vertical distribution is changed from 2.278 to2.857 in the image preprocessing mode and the equilibrium load distribution Times,performance increased by 56.10% to 65.00%.The cutting point and the angle of cutting of the scion and rootstock were extracted from the cutting angle of the grafted seedling,the position of the grafting on the grafted seedling,the cutting point of the rootstocks,the cutting angle of the rootstocks and the cutting point and cutting angle.(3)Research on terminal location and attitude detection method based on monocular vision.Firstly,2 ×2 groups of color concentric circles are designed according to the principle of monocular vision 3D measurement and attitude detection.A posture detection method of iterative query strategy is proposed to realize the detection of landmark plane and terminal attitude.The results of the known concentric circle show that the attitude error method is 0.0469 ° ~ 0.0773 °,and the measurement distance is between 50 and 195 mm.The monocular vision measurement method has the largest Z coordinate of manual measurement The error is 4.87 mm,the minimum error is-0.83 mm,the average error is 1.62 mm,the average error rate is1.31%.(4)Design of grafting cutting device.Mainly around the cutting regulator to achieve the main function,the design of the control system of the scissors seedlings and grafting posture adjustment mechanism of the two end of the actuator,analysis and gives the graft cutting control system of the action process.