基于物联网和图像识别的环境监测系统  

Environmental Monitoring System Based on Internet of Things and Image Recognition

作　　者：陈敏敏[1] 彭安娜 黄道燚[1] CHEN Minmin;PENG Anna;HUANG Daoyi(Shanwei Institute Of Technology,Shanwei 516600,China)

机构地区：[1]汕尾职业技术学院,广东汕尾516600

出　　处：《工业技术创新》2022年第2期112-124,共13页Industrial Technology Innovation

基　　金：2020年度广东省普通高校特色创新项目(2020KTSCX317);2020年广东省科技创新战略专项资金(“攀登计划”专项资金)项目(pdjh2020b1258)。

摘　　要：为满足不同规模环境智能控制需求,开发基于物联网、应用传感器和图像识别的蔬菜大棚环境监测系统,提出了软硬件系统设计方案。上位机基于Android App环境,实现系统显示和用户端的人机交互;下位机由OpenMV图像识别处理模块和三个Arduino芯片构成。OpenMV图像识别处理模块对植物图像进行采集和处理,从图像中获取绿色像素点信息,结合植物的生长天数对植物的生长情况作出分析。三个Arduino(ATmega328)芯片中,一个作为主控芯片,负责配置传感器协议及数据通信;第2个作为控制驱动处理芯片,接收主控芯片的指令,以对外部设备进行控制;第3个采用PID算法,控制调节植物生长的蓝紫灯光,以保持光照度恒定。以小白菜蔬菜大棚为例开展实验,结果表明:在应用该监测系统后,环境温度、土壤湿度与理想值之间仅有2%以内的相对误差,光照度与理想值之间仅有5%以内的相对误差。与同样地理环境下的纯人工监测相比较,可减少约70%的人工劳作量,植物生长时间缩短10%左右。系统实现了环境温度、土壤湿度、光照度等环境参数的采集与控制,具备良好的推广价值。In order to meet the demand of intelligence control with different scales,an environmental monitoring system of vegetable greenhouses based on the Internet of Things technology by using the sensors and image recognition was developed,and the design scheme of software and hardware system was put forward.Based on the Android App environment,the host computer realizes the human-computer interaction between the system display screen and the user terminal;the lower computer consists of an OpenMV image recognition and processing module and 3 Arduino chips.OpenMV image recognition processing module collects and processes plant images,obtains green pixel information from the images,and analyzes the growth of plants according to the growth days of plants.Among the 3 Arduino chips,one is the master chip,which is responsible for configuring the sensor protocol and data communication;the second is as a control drive processing chip,which receives the instructions of the main control chip to control the external devices;the third uses PID algorithm to control and adjust the blue-violet light of plant growth to keep the illumination constant.Taking pakchoi vegetable greenhouse as an example,the results show that after using such a monitoring system,the relative error of environmental temperature and soil humidity against ideal value is only within 2%,and the relative error of illumination against ideal value is only within 5%.Compared with pure artificial cultivation in the same geographical environment,it can reduce about 70%of manual labor,shorten the plant growth time by about 10%.The system realizes the collection and control on the environmental parameters such as environmental temperature,soil humidity and illumination,with a good popularization value.

关 键 词：物联网 环境监测 智能识别 Android App OpenMV 光照度 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术] TN91[电子电信—通信与信息系统]