基于图像识别的水面垃圾智能分类清理装置

Intelligent Sorting and Cleaning Device for Surface Waste Based on Image Recognition

为了有效解决小型水域水面垃圾清理和分类的问题,设计了一种基于图像识别的无人驾驶水面垃圾智能分类清理装置。装置采用平衡式双船体结构,配以翻斗式垃圾收集头及可移动渔网式分类收集仓,并使用涵道式螺旋桨作为驱动;以STM32作为主控,使用OpenMV视觉模块并搭配卷积神经网络算法,实现垃圾类别的识别,并使用ESP32实现对装置的无线通讯与远程监控;基于GPS导航系统、惯性导航系统、视觉导航系统实现自动回航、自动回航等功能;同时,装置还设计有附加的水域参数测量、鱼类投喂等功能模块,以实现装置的一机多用。

In order to effectively solve the problem of water surface waste cleaning and sorting in small waters,an intelligent classification and cleaning device based on image recognition is designed.The device adopts a balancing double hull structure,equipped with a bucket-type garbage collection head and a removable fishing net-type classification collection warehouse,and uses a walkway propeller as the drive.With STM32 as master control,OpenMV vision module and Convolution Neural Network algorithm are used to realize the recognition of garbage category,and ESP32 to realize wireless communication and remote monitoring of devices.Based on GPS navigation system,inertial navigation system,visual navigation system to realize automatic return,automatic return and other functions.At the same time,the device is designed with additional water parameters measurement,fish feeding and other functional modules,to realize the device's one-machine versatility.