| Agricultural Internet of things is the combination of Internet of things technology and agriculture,and the specific application of the various links in agricultural production.Agricultural Internet of things is the key technology to promote modern agriculture from digital agriculture,precision agriculture to wisdom agriculture.Overall sensing and wireless transmission of agricultural information is the starting and key point of development of Agricultural Internet of things technology.Wireless image sensor networks provide an automated and efficient means for overall sensing and wireless transmission of agricultural information.Wireless image sensor networks are a stream of traditional wireless sensor networks,and are the depth development results of wireless sensor networks.They consists of a series of small image sensor nodes in ad hoc network form,and theycan obtain image and video information of the crops and farmland environment,providing intuitive,real-time and convenient means for farmland monitoring,and providing reliable data support for the further image processing in order to obtain the depth information of crops.However,wireless image sensor networksare not massively applied for farmland monitoring.There are some main reasons.Firstly,the performance of the nodes is not up to the needs of users.Secondly,there are not reliable transmission control protocols for the nodes to transfer a large mount of image and video information.Thirdly,the level of management and application of image/video is not high,and users can not obtain satisfactory service quality.So,it is very important to study key technologies of wireless image sensor nodes for farmland monitoring,which is of great significance to promote the development and application of wireless image sensor network technologies in the field of farmland monitoring,to improve the level of agricultural information collection,as well as the development of Agricultural Internet of things technology.In this thesis,we will study key technologies of wireless sensor node,such as design of CMOS image sensor node,design of multispectral image sensor node,data transmission control mechanism of the node,and visual information management platform.The main research contenst are as follows:(1)A wireless image sensor node capable of adjusting its resolution in real time is designed.The node consists of an image acquisition module,a processor module,a wirleless communication module and a power supply module.Based on high-performance and low-power S3C6410 core,the processor module is constructed by the peripheral circuit that is self designed by us,and on the basis of this,embedded Linux is transplanted as the operating system.The low cost image acquisition module is developed by ourselves based on CMOS image sensor chip.A multi-resolution driver control algorithm is designed to ensure the high resolution and adjustment of resolution.Aresolution real-time adjustment algorthim is designed based on cooperaton between application layer and driver layer,and multi-thread concurrency,and it is proved to be validate in experiments.In order to improve the speed of image transmission and expand the communication distance,the method of image transmission based on WiFi technology is studied to replace ZigBee technology commonly used in wireless sensor networks.The solar power supply system is designed to prolong the life cycle of the node.Finally,compared with the current mainstream image sensor nodes,the node shows its superiority.(2)A low cost multispectral image sensor node is designed.At present,the existing multispectral acquisition image devices have some shortcomings,such as expensive in price,huge in volume,low in resolution,and so forth.Further,it is labor expensive and time consumed to carry the device to reality environment.To this end,a low cost multispectral image sensor node based on embedded systems is designed in this paper.A linear mechanical optical filter switching device is developed based on the light splitting principle of optical filter,and a low cost multi spectral imaging system is developed.Optica filter control algorithm is designed to realize the automatic switching of the filters and automatic transmission of multispectral iamges.The spectral image transmission scheme based on 3G is studied,which make the multispectral imaging device has the ability to remotely transfer multispectral images.Through some experiments,the precision of the positioning of the filter,the time overhead of the optical filters switching,the quality of the imaging system were tested,and the multispectral node is compared withmultispectral equipment MS4100 which is widely used in agriculture.(3)The data transmission control mechanism of wireless image sensor node is designed.According to the characteristics of wireless image sensor nodes designed by us,combined with the practical application requirements,a set of transmission control mechanism with hardware and software combination and multi-layer cooperation is designed.The mechanism includes: Classified implementation of communication protocols,multi-function data packet format,multi-level data backup method,adaptive error control scheme for hierarchical classification,clock double synchronization strategy based on DTMS,and so on.(4)According to the characteristics of wireless image sensor nodes,combined with the demand of farmland monitoring,a remote monitoring system for farmland is designed.The topology structure and transmission model of the monitoring system are analyzed,and then the system structure of the monitoring system based on 4G and Wi Fifor large scale collection and remote transmission is proposed.The visual information management software Web-based is designed,which can be used to remotely and visually control wireless image sensor nodes and the network,and effectively manage and use the data collected by the nodes,and provide high quality network service for end users.(5)Wireless image sensor node deployment and comprehensive tests are carried out.According to the application scheme of remote monitoring system,two kinds of wireless image sensor nodes designed by us to be deployed on farmland in a certain scale,and the visual information management software proposed by us is used to control and manage the nodes and the network,and then some comprehensive tests care carried out by using transmission control mechanism.Finally,the test results are analyzed in detail. |
PDF Full Text Request