Research And Realization Of Image Automatic Identification System For Bactrocera Based On SVM

The fruit fly,as a harmful insect with reproductive capacity,wide host range,and seriously affect the quality and yield of fruits of horticultural crops,make the economy suffer huge losses,has also been a quarantine important quarantine object,So it is necessary to know and understand it.Most of the existing automatic identification techniques for insect images are based on the implementation of the PC end.with the swift and violent development of mobile Internet,it has become a necessary product in people's daily life,therefore,developed application software which can automatically identify the species of fly at mobile terminal,can greatly improve the applicability of the application,reduce the output of a large number of human and material resources,simplify the quarantine work,and then promote the import and export trade of related agricultural products.This paper introduces the research background,the research situation at home and abroad,and the technology of automatic identification of fruit flies.and finally,the automatic recognition system of fly image based on Android platform is designed.The functional modules of this system include image acquisition,image preprocessing,automatic image acquisition,classifier design,image recognition and so on.Collected by using the mobile terminal on the fly vein image,The image is preprocessed to improve the quality of the image and reduce the influence of other factors,so as to improve the accuracy of recognition.Using Shi-Tomasi corner detection algorithm detects the image coordinates of the intersection of Drosophila wing,and then use the support vector machine algorithm to classify the fly image,finally through the relevant information on the fly image vein junctions to automatic identification of species of fruit fly.The realization of this system will become the basis of the study of automatic identification of insect images in mobile terminals.