Study On Plant Leaf Recognition Based On Convolutional Neural Network

Plant protection and research needs plant classification and identification,and plant recognition has a certain role in promoting the popularization of plant knowledge and the development of botany related disciplines.Traditional plant identification method usually depends on the plant morphology,texture,color and other characteristics,which can identify which according to plant flowers,fruit,leaf morphology classification work.And,unlike other organs,plant leaf shape is generally tend to flat,easy to collection,survival time is longer,the differences between species,so through the leaves of plant classification recognition has become a hotspot of many scholars.With the development of computer image recognition technology,it is possible to use computer to automatically recognize and classify plant leaves.Convolutional neural network is one of the most popular computer deep learning algorithms.It has been widely used in image processing in recent years.In this paper,a plant leaf recognition model is constructed based on the convolutional neural network algorithm,and 15 kinds of leaf images are selected to test.During the experiment,according to the characteristics of leaf image,to adjust the model parameters to optimize the structure of the model,and the use of self-made leaf image database as the training sample data,complete testing recognition rate.Relying on the advantages of convolutional neural network self-learning,it is not necessary to deliberately intervene in advance,can better get the characteristics.From the experimental results,the recognition model of leaf recognition rate is higher,the average recognition rate can reach 93.7%.In order to expand the scope of application and reduce the use threshold,this paper tries to combine the Android system and the convolutional neural network,and develops a blade intelligent recognition application program based on the Android operating platform.The application includes image acquisition,image processing,recognition,result display and other functions.The blade processing procedures,to form a plant leaf recognition system.The recognition of plant images has good recognition effect,improves the recognition accuracy,availability,and has certain research significance and practical value.