Research And Design Of Virus-based Mushroom Identification System Based On Android

At present,identification methods of poisonous mushrooms mainly include morphological identification based on folk experience,chemical analysis and animal testing.In the actual detection of the above identification methods,there are some deficiencies,such as the need for complex experimental instruments,unsatisfactory detection of unidentified toxins,low accuracy and long required experimental cycle.Poisonous mushroom identification method of this article in view of the present existing problems,on the Android platform combined with image processing and pattern recognition technology,network communication,database technology and other knowledge,study design a poisonous mushroom auxiliary recognition system based on Android mobile terminal,in order to realize the real-time fast identification of poisonous mushroom can be classified,raise their awareness of poisonous mushrooms.The system adopts C/S architecture,including client and server.In this paper,wild mushrooms in hulun buir region of Inner Mongolia autonomous region are taken as the research object.In MATLAB software environment,the collected mushroom images are cut and segmentation.According to their macroscopic features,color and texture feature vectors are extracted,and the combination of feature vectors is constructed as the input of classifier.A prior classification model of poisonous mushroom based on BP neural network was established according to the nonlinearity of extracted characteristic parameters.In the Android platform for the client's development,the main implementation of image acquisition,tailoring and transmission function,the user to enter the client need to choose to stay after identifying the image,or by calling the client camera select add in the local photo album,after cutting the choice of image classification and uploaded to the server for recognition.After the server receives the image,it calls the poison mushroom classification model to identify and classify it,and returns the recognition result to the client.Finally,TCP/IP communication protocol is used to complete the communication between client and server.The developed system was tested,and the experimental results verified the feasibility of the algorithm.The android-based toadstool identification systemcan effectively classify toadstool in real time.