| Our country needs to store large amounts of grain every year,but stored grain insects will cause a lot of losses to grain in the process of grain storage.Therefore,mastering the occurrence of stored grain insects in grain depots is a very important research topic.Detection of stored grain insects including many kinds of methods,this paper uses the method of image recognition to detect stored grain insects in grain depot,according to the stored grain insect images shooting by insect-trapping device,we designed a deep neural network for small target detection,realize the online detection system for stored grain insects,the main research work of this paper is as follows:1.Adults of following six species of common stored-grain insects mixed with grain and dockage were artificially added into the developed insect-trapping device:Cryptoleste Pusillus,Sitophilus Oryzae,Oryzaephilus Surinamensis,Tribolium Confusum,Rhizopertha Dominica.Database of Red Green and Blue(RGB)images of these live insects was established.Color jittering and other methods were used to augmented the database.2.Image processing and SVM classifier were used to detect grain insects without impurities.The segmentation of stored grain insects is realized by watershed algorithm,and then the divided image is used to train the SVM classifier to classify the six kinds of stored grain insects.However,this method could not detect the image containing impurities.3.A detection and identification method for stored-grain insects was developed by applying deep neural network.We used Faster R-CNN to extract areas which might contain the insects in these images and classify the insects in these areas.An improved inception network was developed to extract feature maps.Excellent results for the detection and classification of these insects were achieved.The test results showed that the developed method could detect and identify insects under stored grain condition,and its mean Average Precision(mAP)reached 88. |
PDF Full Text Request