Endangered Wildlife Monitoring And Identification System Based On Image Recognition Technology

The theft and hunting of wild animals by poachers,coupled with the expansion of human activity space and the destruction of the ecological environment,have caused serious harm to the survival and reproduction of wild animals,causing many wild animals to be endangered or have become extinct.Wild animals not only have extremely high research value for our scientific research but more importantly,they are an indispensable and important part of natural resources.Their survival and development maintain the balance and stability of the entire natural ecosystem and affect our human society.Development and survival.By monitoring and identifying wild animals,it can play a very good role in protection.However,traditional detection systems have problems such as difficulty in monitoring protected animals and low accuracy.Therefore,this paper uses the Py Torch framework as the basis and adopts an improved convolutional neural network SSD algorithm to establish a monitoring and recognition system with five animals,the golden monkey,peacock,giant panda,tiger,and elephant,as the research objects.Firstly,an animal image sample library is established by collecting pictures.The pictures in the sample library are labeled with the three-dimensional labeling software labeling,and the training set and the sample set are divided according to the ratio of 7:3.Then invest in the neural network model that introduces the attention mechanism for training,and introduce migration learning and data expansion methods in the training,and finally put the pictures collected by the hardware platform into the network model for positioning and recognition.Through the test of the data set,it can be found that the recognition accuracy rate has reached 96.6%,and the detection speed and the model size are in line with expectations.Besides,the platform is based on Labview to create a multifunctional platform that includes a login interface,wild animal species display,common algorithm recognition,and improved algorithm recognition.Realized the interface display function.Based on the development and research of the above technical functions,this system realizes the image monitoring and recognition functions of five endangered wild animals: the hardware platform collects and transmits images of the animals,and the recognition system uses improved and optimized algorithms to locate and identify the captured images.Recognize and display the recognition results on the software platform,and purposefully save the required images.On the one hand,the system can provide real-time monitoring,which can provide better conditions for the protection of endangered wild animals;at the same time,it saves human resources and improves work efficiency,and enhances the protection of endangered wild animals.