Design And Experiment Of Machine Vision Based Tray-type Green Onion Seedling Machine Complementary Seeding Device

The cultivation of green onion(Allium fistulosum L.)is a traditional advantage and characteristic industry in China’s economic crop production.Industrial seedling cultivation is an effective means to improve the efficiency and quality of onion seedlings,but under some uncontrollable conditions,such as vibration,pressure changes,and so on,it is easy to have a few missed seeds and empty holes,which seriously restricts the seedling quality of tray-type green onion seedling machine and reduces the operational efficiency.To solve this problem,this paper proposes a majority mechanism-based method for identifying miss-seeding tray holes using image processing methods.A test bed of The machine vision-based complementary seeding device of tray-type green onion seedling machine was built to test and verify the recognition and complementary seeding effect.The main research contents are summarized as follows:(1)This paper proposed a miss-seeding tray hole recognition method based on the majority mechanism.For the image acquisition environment of continuous motion of the tray,the image processing of the tray image is performed simultaneously in three different color spaces,including color thresholding,morphological processing,H-Dome method and image reconstruction,and then the majority mechanism voting of the processing results in the three spaces determines the coordinates of the location of the miss-seeding tray holes,and thus realizes the real-time recognition of the miss-seeding tray holes under the high-speed moving conditions of the tray.(2)This paper proposed an machine-vision-based complementary seeding device for traytype green onion seedling machine.According to the control strategy and demand of the pneumatic complementary seeding unit,the research designs the seeding control PCB as the control core,and realizes the coordinated control of the components of the pneumatic complementary seeding unit.Designing and optimizing the structure and parameters of the vacuum generating nozzle by numerical simulation and orthogonal test in combination with the principle of vacuum generation;The liftable sliding seed case was designed in combination with the parallelogram mechanism,and experiments were conducted to optimize the morphology of the seed storage nests and the working time of the case;and the nozzle array and liftable sliding seed case were produced by 3D printing technology.A data sending mechanism between image recognition and the seed complementary unit was established to determine the nozzle pickup and placement time to ensure the reliability of software and hardware interaction and the accuracy of pneumatic seed complementary.This paper built the overall circuit and the overall air circuit of the pneumatic complementary seeding unit to ensure that the complementary seeding unit can receive accurate control.(3)A machine vision-based complementary seeding device for tray-type green onion seedling machine test platform was built to verify the identification effect of the miss-seeding tray hole identification method and the complementary effect of the pneumatic complementary seeding unit.The results showed that the average precision rate of miss-seeding hole recognition was 98.48%,the average recall rate was 97.00%,and the average F1 value was97.73%.The miss-seeding rate of the tray-type green onion seedling machine after installing the complementary seeding device test bed was 0.890%,which was 4.484% lower than before the installation.The machine-vision-based complementary seeding device for tray-type green onion seedling machine designed in this paper achieved the expected requirements of the project,realizing accurate recognition and supplementation of miss-seeding tray holes,and provides some technical support for the practical application of green onion factory seedling complementary seeding devices.