Design And Experiment Of Garlic Seed Orientation Device Based On Image Recognition Technology

Garlic is one of the most important cash crops in China.Single grain sowing in garlic planting needs to follow the agronomic requirements of the upward bulbil and downward root.Since foreign garlic planters cannot be used for hybrid garlic sowing in our case,it becomes necessary for domestic garlic planters to adjust the direction of various garlic bulbils using mechanical devices.These garlic planters have a good effect on the sowing of Lanling garlic,but not good on that of the hybrid garlic.The reason is that the irregular shape of the hybrid garlic cannot meet the condition of the center of gravity locating at the lower half of the garlic clove.In this case of the irregular garlic,the mechanical directing device also fail to adjust the direction of the garlic bulbil.Aiming at the problem that it is difficult for the existing garlic planters to straighten Jinxiang hybrid garlic scales with pure mechanical structure,the head and tail characteristics of garlic were studied.A garlic seed oriented test bench was designed,using image processing technology,according to the different head and tail characteristics of garlic,to identify the direction of garlic bulbs,and to correct the garlic.The research work was mainly carried out as follow:(1)Study on head and tail characteristics of garlic seeds and design garlic bulb identification device.Jinxiang hybrid garlic seed was selected as the research object.The results showed that: Under the same conditions,the energy of garlic head is greater than that of tail,the entropy of garlic head is less than that of tail,and the moment of inertia of the garlic seed head is smaller than that of the garlic seed tail,the texture characteristics of heads and tails of garlic seeds were significantly different.A garlic bulb recognition device is designed to identify the garlic bulb direction based on different texture features of the garlic head and tail.(2)Research on the horizontal arrangement device of Garlic Seeds.According to the scale identification requirements of garlic seed oriented test bench,the horizontal arrangement device of garlic seeds was designed,and the discrete element method(DEM)software was used to simulate the experimental analysis of the horizontal arrangement device of garlic seeds,and the factors affecting the horizontal arrangement of garlic seeds were analyzed.According to the analysis results,the structural parameters of the horizontal arrangement device of garlic seeds were optimized.(3)Research on intelligent lever-type garlic bulb centering device.A lever-type garlic seed bulb centering device was designed,and the discrete element method(DEM)software was used to simulate the experimental analysis of the lever-type garlic bulb centering device of garlic seeds,and the factors affecting the righting of garlic seeds were analyzed.According to the analysis results,the structural parameters of the lever-type garlic bulb centering device were optimized.(4)Research on Intelligent Control System.An automatic control system was designed,taking the Raspberry Pi as the core,using image recognition technology to identify the direction of garlic bulbs,and leveraging the garlic seed centering mechanism to correct the direction of garlic bulbs.The head and tail recognition system of garlic seeds was tested,and the related problems existing in the test were optimized and improved.(5)The garlic seed oriented test bench was tested,and the reasonable working parameter range was obtained through the test.Taking the motor speed,channel width,shooting distance,and recognition threshold as experimental factors,the garlic seed bud righting rate is an experimental index,and a four-factor three-level quadratic regression orthogonal test was performed.Based on the experimental results,a mathematical model of response surface was established,and the influence of each experiment on the recognition rate of garlic bulbs was comprehensively analyzed.The analysis of the interaction factors showed that the influences on the righting rate of garlic bulb scales in descending order were channel width,motor speed,shooting distance,and recognition threshold.The experimental factors have been comprehensively optimized.The optimal working parameter combination is: motor speed is 18 r / min,garlic channel width is 38 mm,shooting distance is 8.6 mm,and recognition threshold is 178,the garlic buds righting rate was 90.56%,and the average identification time of each garlic seed was 0.29 s,which satisfies the requirements of garlic seeders.This research can provide a reference for solving the problem of upright seeding of hybrid garlic.