The Design And Implementation Of Stored Grain Pests Detection System

With the development of science and technology,grain production has been greatly improved,the quantity of grain stored are also increasing,but at the same time,grain losses caused by grain insects are also increasing.The detection and prevention of grain pests is the research focus in the process of grain storage,it is very important to accurately and timely detect and control the stored grain pests effectively.However,at present,most of the grain pests detection methods in domestic grain depot are mainly manual operation,which is cumbersome,heavy workload and difficult to realize automation and informatization.Although there are some scholars to design intelligent embedded type grain insect image acquisition device,but grain pests will set the device accumulated on board,and the collected images do not make any processing,contains a large amount of redundant information;In addition,there is not a complete set of defects in the field of food worm detection,such as image collection,transmission and field identification.To solve these problems,we design a stored grain pests detection system,the system consists of image acquisition subsystem,relay transmission subsystem,data processing subsystem,the main work in this thesis is summarized as follows:(1).Aiming at the problem of high redundancy information in most embedded grain pests image devices,by subtracting the redundant parts from the background subtraction method,the effective grain pests information is extracted,and the network transmission data is reduced,and the system efficiency is improved.The sensor group,which is composed of three photoelectric sensors,can detect the signal of pests,and can provide precise collection instructions even when the processor is in hibernation state,thus reducing the power consumption of the system as much as possible.Every time after each collection control linked to set pests plate stepper motor rotate 180°,to prevent the accumulation of grain pests on the set of worm board,which affect the next collection.(2).Aiming at the problem of same frequency interference in wireless network,the wireless channel division mechanism is designed in the image acquisition subsystem,which effectively avoids interference.Aiming at the problem of short distance wireless communication,combined with the granary site layout,design the relay transmission subsystem network topology,multi-radio multi-channel distribution scheme,on the basis of the extended communication distance,solved the problem of data transmission conflict in the cross-communication area.Finally,based on in-depth analysis of the characteristics of the radio frequency module,designs the low-power receiving scheme based on sleep patterns,launch power control scheme based on received signal strength indicator,on the basis of ensuring the accuracy of data transmission,the energy consumption of the relay transmission subsystem is greatly reduced.(3).Aiming at the problem of useing cloud computing applications creates additional overhead in grain detection system,the data processing subsystem is equipped with a Python experimental platform on an embedded ARM + linux system.The platform first contacts the received RGB565 data stream by BMP coding and then use binarization to segment the image,according to the image characteristics,the region descriptor,invariant moment and texture feature information are extracted.At last,the model of pest species identification based on neural network is constructed.The experimental results show that the success rate of model recognition is up to 93.3%,Meet the actual application needs.The test results show that this design has a stable function and low power consumption,solve the current problems in the field of grain pests detection,it has the strong application and promotion value.