Crop Growth Recognition Monitoring System Based On Convolutional Neural Network

With the advancement of agricultural modernization and the continuous development of information technology,the transformation of traditional agriculture to digital information agriculture has become a systematic project to maintain sustainable agricultural development.The application of modern Internet of Things technology in agriculture can greatly promote the adjustment of industrial structure and the improvement of production quality.Therefore,real-time monitoring of the elements of the farmland environment and accurate acquisition of crop growth conditions to achieve the technical landing of AIOT in agricultural applications are urgent needs of modern smart agriculture.Aiming at the problems of low efficiency,high cost,and low monitoring accuracy of traditional agricultural monitoring methods,and combined with the actual needs of smart agriculture,this paper designs a crop growth recognition monitoring system based on convolutional neural network.This system adopts the modular design concept,has strong scalability,and adopts multi-task division and scheduling based on operating system with strong stability.The upper computer software platform is rich in functions,integrating data communication,image recognition,and acquisition control.The lightweight convolutional neural network Mobilenet is used to identify the corn growth cycle,which has a high recognition rate and fast recognition speed,and the calculation amount and memory size are greatly reduced compared to the standard convolutional neural network.The specific research content is as follows:The crop information monitoring device is mainly responsible for the collection and transmission of image and environmental positioning data.The system uses solar energy to supply power in the wild,and supplies different voltages to the system through charge and discharge hardware circuits and voltage conversion circuits.The software platform is written based on the freertos operating system,and reasonable tasks are divided based on the modular programming idea to ensure that the system runs in an orderly manner.The host computer software uses a convolutional neural network to identify and analyze the growth stage of the collected corn growth images,and display the received environmental positioning data and save it in the database for later viewing and data analysis.The uppercomputer software platform is a comprehensive platform that integrates functions such as data communication,image processing,information storage,interface display and acquisition control.The convolutional neural network was used to identify the growth stage of the collected corn image.Train the standard convolutional neural network VGG16 and the lightweight convolutional neural network Mobilenet separately,input the same test set into two network models,compare the network performance,and finally use a Mobilenet convolutional neural network that has both a high recognition rate and a small computation.Field tests show that the system's crop growth identification monitoring device and the host computer software run smoothly and efficiently in cooperation,crop growth images and environmental positioning data are collected quickly,data processing and crop growth identification have high stability and accuracy,which fully meets the actual application requirements.Therefore,it is of great significance and application value to combine the Internet of Things model and convolutional neural network in agricultural production.