發(fā)布時間：2018-06-26 11:32

  本文選題：機(jī)載InSAR + DEM��； 參考：《測繪科學(xué)》2017年06期

【摘要】：針對如何提高機(jī)載InSAR DEM空洞填補(bǔ)精度和DEM對地形細(xì)節(jié)描述的難題,該文對分形模擬的中點位移插值函數(shù)做了改進(jìn)。首先,提出了動態(tài)分形參數(shù)提取思路,根據(jù)選定的窗口動態(tài)提取局部分形參數(shù),使提取的分形參數(shù)局部精度顯著提高。其次,提出了分形曲面中心內(nèi)插法,以正方形為圖像基元,將兩點信息擴(kuò)展到8點信息來充分顯示局部變化的細(xì)節(jié)信息;考慮到對采樣點點數(shù)和位置的要求,結(jié)合距離加權(quán)平均法來彌補(bǔ)周圍不夠8點的特殊情況。最后,以國產(chǎn)機(jī)載X波段全極化InSAR影像提取的DEM為數(shù)據(jù)源進(jìn)行了實驗。實驗結(jié)果表明,相比傳統(tǒng)的中點分形內(nèi)插法,該方法不僅能有效提高空洞填補(bǔ)精度,且能更好地描述地形的細(xì)節(jié)信息。
[Abstract]:In order to improve the accuracy of filling holes in airborne InSAR Dem and the difficulties of Dem in describing terrain details, the mid-point displacement interpolation function of fractal simulation is improved in this paper. Firstly, the idea of dynamic fractal parameter extraction is put forward, and the local fractal parameter is extracted dynamically according to the selected window. The local precision of the extracted fractal parameter is improved significantly. Secondly, a fractal surface center interpolation method is proposed, in which the square is taken as the image primitive, and the two points information is extended to 8 points to fully display the local variation details, and the requirements for the number and position of the sampling points are taken into account. The distance weighted average method is used to compensate for the special case of less than 8 points in the surrounding area. Finally, Dem extracted from domestic airborne X-band fully polarized InSAR images is used as data source. The experimental results show that compared with the traditional midpoint fractal interpolation method, the proposed method can not only improve the accuracy of cavity filling, but also better describe the detailed information of terrain.
【作者單位】： 中國測繪科學(xué)研究院;中南大學(xué)地球科學(xué)與信息物理學(xué)院;
【分類號】：P208;P237
，

本文編號：2070251