基于PaddlePaddle深度学习的野外露营安全装置

A Field Camping Safety Device Based on PaddlePaddle Deep Learning

针对近年野外露营存在的环境及安全隐患等问题,设计一种基于PaddlePaddle深度学习的野外露营安全装置。首先采用卷积神经网络训练模型,然后通过图像识别分类技术进行识别。文中以动物分类识别为例,通过对不同种类动物图片进行训练来改进模型,并对模型识别及分类的准确性进行验证,最终精度为0.96以上。经测试,采用的模型在动物图像分类中有较好效果,能满足课题设计要求。

Aiming at the problems of environment and safety hazards in field camping in recent years,a field camping safety device based on PaddlePaddle Deep Learning(DL)is designed.Firstly,a Convolutional Neural Networks is used to train the model,and then image recognition and classification techniques are used for recognition.Taking animal classification and recognition as an example in the paper,the model is improved by training images of different types of animals,and the accuracy of model recognition and classification is verified.The final accuracy is above 0.96.After testing,the model used has shown good performance in animal image classification and can meet the requirements of the project design.