Design Of Embedded Wireless Image Detection System For Grain Pests

The detection methods of pests in stored grain cover artificial recognition,voice recognition, near infrared spectrum and image recognition method. Now themain frequently-used detection methods of pests in stored grain are visual detection,sampling inspection and human temptation. But these methods mostly depend onmanual operation, which are belonging to the artificial recognition method with thedisadvantages of fussy operation, heavy workload, difficulty to realize automationand informatization. Detecting the pests in stored grain by using computer imagerecognition technique has the advantage of high precision, small amount of labour,efficient and easily connecting to the existing computer management system. Butthe problem of data sources of the pests in stored grain is not solved.This paper is mainly focus on solving the above problem by designing anEmbedded Wireless Image Detection System for Grain Pests. This system is used torealize the automation and intelligent of pest’s image acquisition in small andmedium size grain depot by using wireless communication on Sub-GHz, embeddedsystem and TCP/IP. Image acquisition node in the system be made up of microcamera, CPU, radio-frequency circuit, LED illuminant, button battery, pests’accumulator and so on, which can be put the place that stipulated by the TechnicalCriterion for Grain and Oil-seeds Storage, the place usually have a certain depth inthe grain bulk. The pets in stored grain can be attracted to enter into the pests’accumulator by bait and LED illuminant put in the image acquisition node. CPUcontrol the camera to capture the pests image while receiving instructions fromconsole cabinet, the send the image that processed simply to console cabinet byimage forwarding node. It realizes the image acquisition of pests in granary forimage recognition of pests and providing the data sources of estimating the level ofInsect pests.This paper has carried out and accomplished these works as below:1. Solution design of the Embedded Wireless Image Detection System forGrain Pests.2. Software and hardware design of image acquisition node and imageforwarding node.3. The migration of the real-time kernel μC/OS-III and the embedded protocolstack μC/TCP-IP based on STM32F107.4. Functional test of the image acquisition node and image forwarding node.