Research On Embedded Intelligent Visual Monitoring System For Agriculture

Plant is an important food source for human beings and animals,and also is an important component of maintaining ecological environment healthy and food chain stability.In order to produce high quality plants,farmers work hard in the fields every day,which requires not only a lot of time,but also a lot of labor.Aiming at this problem,this paper designs an embedded agricultural intelligent vision monitoring system,which can be used to identify the types of crops and flowers in greenhouses,to understand their growth status,to keep the plant growing healthily in the most suitable environment and not affected by the seasonal changes and environmental changes,and help workers grow crops more efficiently.This paper uses the camera to collect and monitor the plant in real time,and the plant growth state information is transmitted to the processor for storage and preprocessing.In order to improve the blade recognition rate and recognition speed,this paper combines the clonal selection algorithm and support vector machine algorithm to train the blade,and establish the blade recognition classifier,which identified results and plant growth video can be viewed through the host computer interface.In order to make users know real-time environmental information,the article uses a variety of sensors to collect greenhouse temperature,humidity,light intensity and CO₂ concentration data,which acquired in the greenhouse in realtime is transmitted to the system processor in the manner of ZigBee wireless communication,and users achieve the purpose of sending and receiving commands by GPRS wireless network technology to view environmental parameters and regulate lighting,water pumps,exhaust devices and other purposes.If the environment factor is not in the pre-set threshold range,the system will automatically send alarm messages to the users.While the core processor uses the Linux operating system and S3C6410 processor,using HTML web pages,Boa server,CGI technology to make the images and environmental parameters received in PC display page to realize the growingview and management of remote greenhouse plants.The system is intelligent,high recognition rate,small cost,easy to manage and other characteristics,that can effectively help the agricultural staff to monitor greenhouse crop growth.The identification of plant species facilitate the farmers fertilization spraying,improve the utilization rate of pesticide,reduce the daily workload and labor,increase the yield,the greenhouse planting precision,automation,digital management.This system is intelligent,high recognition rate,low cost and easy management features,that can help the growth of agricultural crops in the greenhouse monitoring staff,through identification of plant species to facilitate the farmers fertilization spraying,improve the utilization rate of pesticide,reduce the daily workload and labor,increase the cultivated land the yield,and achieve cultivation,precision,automation,digital management to agreenhouse.