Design And Development Of Royal Jelly Strip Transfer Larva System Based On Machine Vision

The production of royal jelly includes four steps: cutting beeswax,removing old insects,scraping royal jelly,and transferring larvae.The transfer of larvae is to move the queen bee larvae from the honeycomb to the royal jelly strip,The automated production of transferred larvae is the most difficult,because the larvae are small and fragile,and currently it can only be produced manually.This paper designs a royal jelly strips automatic larva transfer system based on machine vision.The main research contents include:1、Mechanical design: Analyzed the ideal trajectory of manual transfer of larvae.The manipulator for transferring larvae is designed according to the ideal trajectory.Designed a three-axis movement transmission mechanism for the manipulator to transfer larvae,designed the feeding and discharging mechanism of royal jelly strip,complete the actual system construction of the mechanical structure for the transfer of larvae.2、Control circuit and program design: The selection of system controllers,sensors and drive components has been completed.The control circuit for transferring larvae has been designed,the control process has been sorted out,and the control program has been compiled.3、Visual processing system design: A number of images of bee larvae were collected and the image characteristics of the larvae were analyzed.Preprocess the captured images.The larvae are identified and detected by template matching,feature point detection,color recognition,and convolutional neural network algorithms.Compare the advantages and disadvantages of different image detection algorithms in bee larval recognition.Finally choose template matching based on color recognition.And the results show that the algorithm has an accuracy of more than 90% in larval detection.The system of royal jelly transfer larvae transfer based on machine vision has been tested.The results show that the transfer can be successfully achieved,and the transfer success rate reaches 90%,which meets the design requirements.