Research On Visual Agriculture Monitoring System

With the rapid development of science and technology,intelligent agricultural equipment is gradually replacing traditional agricultural planting.Most of the intelligent agricultural monitoring equipment uses fixed-position monitoring and sensors,which cannot dynamically monitor the environmental information of the entire greenhouse.Global large-area multi-span greenhouses cannot be managed separately in blocks when mixing crops;the current intelligent agriculture The research on monitoring equipment aims to help growers improve production efficiency,but it ignores consumers with strong purchasing power.In view of the above problems,this paper designs a visual agricultural monitoring system,integrates sensors and cameras on a movable trolley,and can collect the environmental information of the greenhouse without dead corners through the guide rails set above the greenhouse,and uses image recognition technology to the mixed crop area Management,and provide consumers with a traceability platform.The centralized control module of this system adopts S3C2440 based on LINUX platform as the core processor,and transmits the environmental information collected by various sensors of the movable monitoring module to the server module through WIFI.The server module is built in the host computer,and the data storage is carried out by transplanting the Tomcat server and SQLite database.After preprocessing the video information,the DBN-integrated model classifier is trained by integrated learning,and the plants are identified by the voting method Classification improves the blade recognition rate.Users can monitor the environmental information in the greenhouse by using the Java Web dynamic web page developed by Eclipse,and open real-time monitoring,product traceability and query of historical environmental data functions to consumers through permission settings.Administrator users add remote control and user management functions.If the environmental information in the greenhouse exceeds a preset threshold,the system will control the corresponding equipment of the equipment control module to carry out regulation work.The system adopts mobile monitoring,which can collect environmental information without dead corners.The plant leaf recognition algorithm has high accuracy and strong real-time monitoring video,which reduces the labor of growers,facilitates unified monitoring and management of mixed crops,and improves land use.Rate and production efficiency.Growers can use this system to accurately control the growth environment of crops.Consumers can also use this system to trace the source of crop growth.They can purchase quality products with confidence and have the advantages of brand promotion.