摘要
鸟类是生态系统中的重要组成部分,传统的鸟类调查与监测需耗费大量的人力物力,效率十分低下。对此,基于微信小程序,利用Flask框架和深度神经网络,研究设计了一个鸟鸣声识别系统,通过声音识别技术实现了鸟鸣声的自动识别和分类。用户可通过微信小程序录制或上传鸟鸣声音频文件,系统将自动对音频进行预处理和特征提取,然后使用训练好的CNN模型对鸟鸣声进行识别。同时系统还提供了鸟类字典、数据监测可视化等相关信息的查询和展示。
Birds play a crucial role in ecosystems,but the traditional ways of studying and monitoring birds are inefficient.In order to solve this problem,this paper develops a birdsong recognition system integrated into a WeChat mini⁃program using the Flask framework and deep neural network.The system employs sound recognition technology to identify birdsongs automatically.Users can record or upload the bird song audio files through the mini⁃program.The system then automatically preprocesses and ex⁃tracts features from the audio,and use a trained CNN model for the recognition of bird songs.In addition,the system provides func⁃tionalities like a bird dictionary and visualization tools for data monitoring for the users to look up or display.
作者
陈凌芳
周雁
王庆娟
林佳皓
谌业恒
Chen Lingfang;Zhou Yan;Wang Qingjuan;Lin Jiahao;Chen Yeheng(Beijing Institute of Technology Zhuhai,Zhuhai 519088,China;Beijing Institute of Technology,Beijing 100081)
出处
《现代计算机》
2024年第4期75-82,共8页
Modern Computer
基金
广东省普通高校特色创新项日(2022KTSCX202)
广东省科技创新专项资金(“攀登计划”专项资金,pdjh2022b0711)
北京理工大学珠海学院大学生创新创业训练项目(2023062DCXM)。