Automatic Detection And Control System For Corn Haploid Grain Sorting

Corn being a grain crop with the highest yield in China,the demand for corns is growing rapidly year by year.In actual production,haploid breeding is the one of the most effective breeding methods for the increase of output of corns,which attracts wide attention from breeding workers.To use haploid to breed,the first problem to be solved is to sort haploids out from mixed seeds.However,there is no mature and complete set of automatic equipment for quick sorting of haploid in the current market.Therefore,it is of great importance to study and design an automatic quick sorting device for corn haploid grains,so as to speed up the breeding process and raise output of corns.In this research,an automatic detection and control system for corn haploid grain sorting is designed.The system is composed of two parts of computer-based upper computer platform and PLC-controlled lower computer platform,of which computer and PLC communication are realized by RS232.FX2n-14MR Mitsubishi PLC being the core,sample transmission module,stepping motor module and sample injection module of lower computer are controlled,and then seeds are conveyed to effective acquisition range of image collection module which will send the images to the upper computer for image processing and recognition.Based on recognition results,the upper computer will send the sorting signal to the lower computer which controls airflow sorting module and therefore completes seed sorting.The upper computer software,including man-machine interaction interface,image acquisition software,image processing software and PLC communication software,is designed based on MFC engineering platform of Visual Studio 2010.The lower computer control software is realized by PLC ladder program.In this thesis,MATLAB is utilized to conduct system simulation of sorting and recognition method based on color feature and sorting and recognition method based on texture features.Through comparison,color feature based sorting method is the optimal solution,which is realized by OpenCV.Through hardware debugging of the system,haploid grain sorting is detected and tested.The system test results show that the correct haploid sorting rate is 87.7%and that of diploid is 92.6%,which are similar to the correct haploid sorting rate of 90.2%and correct diploid sorting rate of 91.8%in the system simulation results.Hence,it is verified that the system satisfies the production feasibility and can achieve automatic and correct grain sorting,which offers an effective solution to quick sorting of corn haploid grains.