摘要
针对传统的果实蝇鉴别系统通常采取的是PC端辨识,辨识效率低,不利于移动作业,无法满足实际科研及检疫工作的需求。为提高辨识效率和实用性,开发了基于android平台的果实蝇离线快速辨识系统,运用BP神经网络算法实现对果蝇的分类鉴定。该系统在无网无信号状态下的恶劣野外环境也能够照常使用,使得果蝇识别技术在移动便携设备上的应用成为了现实,解决了传统PC端系统辨识效率低,不利于移动作业的问题。
For the traditional f ruit f lies identif ication system, PC-side identif ication is usually adopted. The recognition eff iciency is low, which is not conducive to mobile operations and can not meet the needs of actual scientif ic research and quarantine work. In order to improve the recognition eff iciency and practicality, the off line fast identification system of fruit f lies was implemented in android platform. The system realizes the collection of fruit f ly images from mobile devices and uses the BP neural network algorithm to achieve the classif ication and identification of fruit f lies. The system can also be used as usual in the harsh field environment without signals and no signal, making the fruit f ly identif ication technology. The application on mobile portable devices has become a reality, which solves the problem of low recognition eff iciency of conventional PC-side systems and is not conducive to mobile operations.
出处
《科技创新导报》
2018年第27期160-161,共2页
Science and Technology Innovation Herald
基金
江西省普通本科高校中青年教师发展计划访问学者专项资金(项目编号:赣教办函[2016]109)
江西省科技支撑计划项目"基于生物电特性的脐橙品质无损检测关键技术研究"(项目编号:20123BBF60177)
关键词
果实蝇识别
移动终端
BP神经网络
Fruit f ly identif ication
Mobile terminal
BP neural network