微波成像中載具動態影響之模擬

Abstract

戰略或民用偵測遙測性載具常因天候狀況、煙塵及夜晚而造成光學系統之失效。微波影像則提供了此問題之一種解決辦法，且其方位分辨率及距離分辨率也趨近類似光學攝影的高分辨率圖像，並可具有更分明的輪廓和對比度，以及更多特殊目標物之細節紋理。微波影像所產生地面影像的品質決定於解析度的尺寸，而解析度尺寸可由距離與方位解析度來明確說明。在距離處理程序中常用匹配濾波器來得到最大的訊噪比訊號輸出並利用線性頻率調變來達到脈波壓縮的效果。而在方位處理程序中即是利用天線本體的運動對待測物體的都卜勒效應來產生線性調變的效果，這需搭配主動式之導航系統及空用電子感測及衛星定位裝置來得到載具之動態資料作為成像流程訊號處理之用。本計畫預計分為兩年期計畫，第一年計畫主要目的是研究微波影像之訊號處理流程，模擬出發射及接收微波訊號至地表上所得到回波合成之影像並利用視覺化軟體來顯示出所模擬之影像。第二年計畫主要目的是模擬飛行載具執行任務之動態過程中所整合載具上之加速規、陀螺儀及全球衛星定位系統之位置資料，對訊號處理流程所影響之微波影像品質之失真性。並透過運動平台與視覺化之軟體將所紀錄之動態資料重現出來，以分類各種載具動態效果對影像之影響關係。

Situations of the weather, dusts, and night cause failure of the optical system on vehicles intention of strategical and civil remote sensing. The microwave images provide a solution of this kind problems. Its range and azimuth resolution are closed to high-quality optical pictures to resolve the contours and the contrast of objects. For some particular buildings, it also have special texture of pattern. The imaging system performance is determined by the pattern pixel size of the resolution cell. Resolution cells are divided into two axes of range and azimuth. Range processing usually emphasizes the signal to noise ratio(SNR) to be maximum. It also utilizes the linear frequency modulation(LFM) to accomplish the pulse compression. Another way, the azimuth processing utilizes the Doppler effects caused by antenna and objects motion to produce the LFM. The data of vehicle dynamic motion are measured by the active inertial navigation system(INS) and aerial sensors for imaging signal processing. The project is divided into 2-years to execute. The purpose of the first year of this project is to research the processes of synthesizing microwave signals and to simulate the transmitted signals and received from the terrain. Furthermore, the simulation patterns also have demonstration by implementing the visualized software. The second year of this project will simulate the vehicles with working dynamic data including integrating the accelerators, the gyroscopes, and the global positioning system(GPS). The procedures of simulation will present the distortion pattern obscured incorrectly by vehicle』s motion and compute out the raw data. The effects of motion will also be reappeared on the platform and classified the images depended on which motion.