航天返回与遥感
航天返迴與遙感
항천반회여요감
SPACECRAFT RECOVERY & REMOTE SENSING
2014年
6期
42-48
,共7页
调制传递函数%数学模型%曲面构造%旋转插补法%像质评价
調製傳遞函數%數學模型%麯麵構造%鏇轉插補法%像質評價
조제전체함수%수학모형%곡면구조%선전삽보법%상질평개
smodulation transfer function%mathematical model%surface construction%rotation interpolation method%image quality evaluation
调制传递函数(modulation transfer function, MTF)是定义在二维频率域空间的实函数,可用一个MTF空间曲面描述,是光学成像系统成像性能的一种重要的评价指标。由于成像系统的二维MTF曲面不能直接测量,需要先测量几个特定方向的一维 MTF曲线,再由曲面构造方法间接获得。对于同样的MTF曲线组,用不同方法构造出的MTF曲面各不相同,因此有必要研究MTF曲面构造方法的有效性。文章对现有构造 MTF 曲面的乘积法和插值法进行分析,提出一种新的旋转插补法,并使用(准)高斯函数和4次多项式函数作为二维MTF曲面的模拟标准,对上述3种方法的有效性和误差分布进行定量分析。实验结果显示,旋转插补法具有更好的适用性。
調製傳遞函數(modulation transfer function, MTF)是定義在二維頻率域空間的實函數,可用一箇MTF空間麯麵描述,是光學成像繫統成像性能的一種重要的評價指標。由于成像繫統的二維MTF麯麵不能直接測量,需要先測量幾箇特定方嚮的一維 MTF麯線,再由麯麵構造方法間接穫得。對于同樣的MTF麯線組,用不同方法構造齣的MTF麯麵各不相同,因此有必要研究MTF麯麵構造方法的有效性。文章對現有構造 MTF 麯麵的乘積法和插值法進行分析,提齣一種新的鏇轉插補法,併使用(準)高斯函數和4次多項式函數作為二維MTF麯麵的模擬標準,對上述3種方法的有效性和誤差分佈進行定量分析。實驗結果顯示,鏇轉插補法具有更好的適用性。
조제전체함수(modulation transfer function, MTF)시정의재이유빈솔역공간적실함수,가용일개MTF공간곡면묘술,시광학성상계통성상성능적일충중요적평개지표。유우성상계통적이유MTF곡면불능직접측량,수요선측량궤개특정방향적일유 MTF곡선,재유곡면구조방법간접획득。대우동양적MTF곡선조,용불동방법구조출적MTF곡면각불상동,인차유필요연구MTF곡면구조방법적유효성。문장대현유구조 MTF 곡면적승적법화삽치법진행분석,제출일충신적선전삽보법,병사용(준)고사함수화4차다항식함수작위이유MTF곡면적모의표준,대상술3충방법적유효성화오차분포진행정량분석。실험결과현시,선전삽보법구유경호적괄용성。
Modulation transfer function (MTF) is a real function defined by the 2-D spatial frequency domain and could be described by a space MTF surface. Although the 2-D Modulation Transfer Function (MTF) surface of imaging system could not be measured directly, it could be constructed by measuring 1-D MTF curves in several specific directions. The 2-D MTF surfaces thus created are not the same, however, this method of construct 2-D MTF surface deserves to ensure its effectiveness. We analyze the current methods of MTF surface construction, product method and interpolation method, and then propose a novel rotation interpolation method. The (quasi) Gaussian function and the 4-degree polynomial function are designed and used as the MTF surface hypothetic standards. By comparing the effectiveness and error distributions between the various methods, the rotation interpolation method is better in applications.