| More and more precise guidance weapons are required in order to win the future war. Terrain-aided navigation(TAN) and scene matching are important technologies in automatic aircraft navigation with high accuracy,but they work with some limit condition and can't meet the demand of the uture war. With the development of new technologies, more and more new sensors are used for navigation. Imaging laser radar(Lidar) is novel kind of distance image sensor with high accuracy by combining the laser technology and radar technology, which work all-weather, day-or-night without being influenced by weather and lighting condition. So it is very important to study on the imaging laser radar terrain matching for automatic aircraft navigation. Therefore, to design the real time information processing machine for the image aided aircraft navigation is one of the key foundations for the radar used in imaging navigation application. In this paper, we analyze the principle of imaging laser radar at first, include the theories for detection model of the Lidar, the method for 3D terrain elevation map recovery, and the arithmetic for image aided aircraft navigation, and so that to determine the demand and task of the information processing machine. Through the transverse optical scanning system and the aircraft's lengthways flying, after trigonometric transform, coordinate conversion and geometry rectification, we can get the terrain elevation map and the reflected Image under the aircraft, then we can fusing the terrain elevation map and the intensity image for the aircraft navigation. Then According to the principle of information processing in LIDAR, the information system is analyzed and the function is divided. Therefore the hardware and software of signal processor machine in LIDAR is designed. As to hardware, the high-performance DSP is adopted as main processor, on the other hand, middle-scale and high-speed FPGA is used to accessorial processor. Multi-level story architecture is used to satisfy the massive image data. The input port of signal processor machine can be connected to the LVDS of LIDAR and USB2.0 of computer which produce synthetic data. The output port is connected to RS422 and RS232 of control cell. The interface of inter-board is designed for extension. As to software, image matching algorithm is proposed using Pyramid multi-resolution in wavelet domain combined with FFT–based fast algorithm. The driver software of signal processor machine and simulation software of computer are also designed. At the end of this chapter, the key technique of design in high-speed digital system is analyzed and the PCB of signal processor machine is designed and tested. Final we discussed the frame work of the software for the LIDAR and also implement the real time arithmetic on the information processor machine. The arithmetic is designed based on multi-feature-based fusing using nerve-network and muti-resolutiong wavelet transform,and have high precision on matching, high performance on real-time and high reliability. The software on the macine can do well the the application. The experimental results show that The architecture is flexible and the signal processor machine can achieve real-time,high reliability. It can satisfy the demand of navigation system. |
PDF Full Text Request