Study On Image Acquisition And Processing System Based On UAV Remote Sensing Platform

Precision agriculture required to master the differences of soil fertility status and crop growth status in time and space of the farmer land, then production personnel can raise filed management strategy based on these differences. Remote sensing is an important means to obtain farmland information. Traditional agricultural means of remote sensing is by satellite and manned aircraft. The high cost and longer period of these two remote sensing methods cannot meet the requirements of the precision agriculture, to some extent they even obstruct the development and popularize of precision agriculture. In recent years, with the expansion of UAV technology in the field of commercial, UAV became a new research focus in the field of agriculture, using UAV as low-altitude remote sensing platform on farmland become a reasonable idea. UAV is flexible, easy to operate, low cost, safe and reliable. It can help to get rid of the limit of weather which is an uncontrollable factor. In this subject I designed a set of image data acquisition, compression and quasi-real-time transmission system, improved the function of the UAV remote sensing platform, the main content of this subject is:(1) Developed a set of ARM processor-based SLR camera control panel.SLR camera control panel is the hardware core of this thesis, it uses the S3C6410chip which bases on ARM11core and uses the mobile power supply of5V2A.The major modules on the panel are:1)Wired network module.The wired network modules can transfer data with the host via the network cable while debugging. Experiment shows this module can work well, the driver of the NIC(network interface card) can immediately identify and establish a connection when the control panel and host is connected by a network cable.2)USB module.The control panel is connected via a USB module with the camera, program from the control panel can send commands to the camera via USB cable and the camera can send image data it collected to the control panel via the USB cable.3)WiFi module.The control panel forms a wireless LAN with ground receivers via the WiFi module,the image data collected by the camera can be transmitted to the ground receivers after compress ion. The practical effects show that the control board is stable and reliable, all modules can work together.(2) Transplanted embedded Linux operation system to the camera control panel. By analyzing the advantages and disadvantages of several commonly used embedded operation system, embedded Linux operation system is the most appropriate, then cross compile and transplant this system to the control panel.Installed a virtual machine of desktop version of Linux-Ubuntu, then downloaded the Linux kernel2.6.36and cross-compiled it in the virtual machine, then write the kernel image file into the control board using the SD card.In order to compile and test the camera control software, the paper chose NFS as a temporary root file system at first.The transplant result shows that the Ubuntu system and the camera control software used in this paper has good compatibility and using NFS as the root file system provides a great convenience for embedded software development.(3) Transplanted camera control software to the control panel.Gphoto2is a digital camera control software developed by some Linux enthusiasts, it follows the GPL and all source code are public, downloaded it from its official website, configured it for the camera control panel, cross-compiled it and then transplanted it to the control panel.Through the Gphoto2software can set the vast majority parameters of the SLR camera.After the UAV takes off the remote sensing control software continuously detect the GPS signal, when the values are same as the default the snapshot commands of the Gphoto2will be called and the image data will be transferred to the memory of the control panel.Experiments show that the Nikon D90SLR camera support the Gphoto2software on the control panel, snapshot commands issued by the control panel received and executed by the camera and the image data can be promptly transmitted to the memory of the control panel.(4) Described two types of image compression algorithm.The compression algorithm includes lossy and lossless compression. As the image data acquired by the Nikon D90is10-20MB, this paper tried to use a classic lossless coding algorithm-Huffman coding method to compress the picture file, the result showed that the effects were not obvious. The paper also discussed a possible high-spectral image compression method for hyperspectral camera that will be used in the future, this is a preparatory work for the next stage.(5) Achieving wireless LAN image data transmission.This paper analyzes several wireless data transmission technology, in consideration of transmission speed, transmission distance and the availability, the wireless WiFi technology is the best one to transmit image data.The wireless NIC of the control panel on the UAV remote sensing platform and the receiver on the ground form a wireless LAN, using socket programming to achieve TCP protocol to transmit data.