發(fā)布時間：2018-06-20 15:05

  本文選題：水體提取 + 黃土高原地區(qū)　； 參考：《國土資源遙感》2017年04期

【摘要】：黃土高原地區(qū)水體特征復(fù)雜多樣,基于高空間分辨率遙感影像在大面積范圍內(nèi),現(xiàn)有提取方法很難兼顧精度與效率,故提出一種新型的水體精細(xì)化自動提取方法。首先,在全局范圍內(nèi)根據(jù)不同的水體類型使用面向?qū)ο蟮挠跋穹治黾夹g(shù)按照不同的規(guī)則方法初步提取水體的主體部分,形成水體的種子區(qū)域;然后,在局部范圍內(nèi)根據(jù)同一地物的光譜相似性,進(jìn)行區(qū)域種子生長,實(shí)現(xiàn)水體的高精度提取。結(jié)果表明,使用該方法提取的水體邊緣可以較好地與影像吻合,更能保證水體的完整性,有效減少干擾信息,提升運(yùn)算效率。
[Abstract]:The characteristics of water body in Loess Plateau are complex and diverse. Based on high spatial resolution remote sensing image in a large area, it is difficult for existing extraction methods to take into account accuracy and efficiency. Therefore, a new method of water body refinement and automatic extraction is proposed. First of all, according to different water body types, the object oriented image analysis technique is used in the global scope to extract the body parts of the water body according to different rules, so as to form the seed area of the water body. According to the spectral similarity of the same ground object, the regional seed growth is carried out in the local scope, and the high precision extraction of water body is realized. The results show that the edge of the water body extracted by this method is in good agreement with the image, which can ensure the integrity of the water body, effectively reduce the interference information, and improve the efficiency of operation.
【作者單位】： 北京地拓科技發(fā)展有限公司;北京林業(yè)大學(xué);水利部水土保持監(jiān)測中心;
【基金】：高分水利遙感應(yīng)用示范系統(tǒng)(一期)“高分辨率對地觀測系統(tǒng)重大專項”(編號:08-Y30B07-9001-13/15) 水利部綜合事業(yè)局拔尖人才培養(yǎng)專項共同資助
【分類號】：TP751;X87

【相似文獻(xiàn)】

相關(guān)期刊論文 前1條

1 孫虎，惠振德，，陳鵬;黃土高原地區(qū)的環(huán)境資源與持續(xù)發(fā)展[J];陜西師大學(xué)報(自然科學(xué)版);1995年03期

本文編號：2044670