本文選題：建筑物檢測 + 圖像分割��； 參考：《數(shù)據(jù)采集與處理》2017年02期

【摘要】：在遙感圖像中,災(zāi)區(qū)建筑物的檢測對災(zāi)情獲取和災(zāi)后應(yīng)急救援具有重要意義。針對災(zāi)區(qū)高分辨率遙感圖像中建筑物檢測的問題,提出了一種改進(jìn)的基于形態(tài)學(xué)特征的多方向和多尺度分割方法,以實(shí)現(xiàn)災(zāi)區(qū)建筑物的自動化檢測。首先將形態(tài)學(xué)算子的重建、粒度和方向等性質(zhì)整合到建筑物的亮度、大小和對比度等特征中,對遙感圖像進(jìn)行初步的分割并提取高亮和高對比度的建筑物,然后結(jié)合圖像的區(qū)域邊緣信息,進(jìn)一步提取潛在的建筑物。實(shí)驗(yàn)結(jié)果表明,所提方法對災(zāi)區(qū)高分辨率圖像中的建筑目標(biāo)有較高的檢測率和較低的誤檢率。
[Abstract]:In remote sensing images, the detection of buildings in disaster areas is of great significance to disaster situation acquisition and emergency rescue. In order to solve the problem of building detection in high resolution remote sensing images of disaster areas, an improved multi-direction and multi-scale segmentation method based on morphological features is proposed to realize the automatic detection of buildings in disaster-stricken areas. Firstly, the reconstruction, granularity and direction of the morphological operator are integrated into the luminance, size and contrast of the building to segment the remote sensing image and extract the high-contrast and high-contrast buildings. Then combined with the region edge information of the image, the potential buildings are further extracted. The experimental results show that the proposed method has higher detection rate and lower false detection rate for building targets in high resolution images of disaster areas.
【作者單位】： 南京理工大學(xué)計算機(jī)科學(xué)與工程學(xué)院;
【基金】：國家自然科學(xué)基金(61003108/61371168)資助項(xiàng)目 公安部應(yīng)用創(chuàng)新計劃(2013YYCXGASS097)資助項(xiàng)目
【分類號】：TP751

【相似文獻(xiàn)】

中國期刊全文數(shù)據(jù)庫 前3條

1 Ejaz Hussain;Serkan Ural;KyoHyouk Kim;Chiung-Shiuan Fu;Jie Shan;劉明眾;;基于2010年太子港震后GeoEye-1影像、激光雷達(dá)數(shù)據(jù)的建筑物提取和瓦礫制圖[J];地殼構(gòu)造與地殼應(yīng)力;2012年01期

2 馮青松;雷曉燕;伍明輝;;地鐵運(yùn)行列車引起建筑物低頻振動的數(shù)值分析[J];鐵道科學(xué)與工程學(xué)報;2007年05期

3 ;[J];;年期

，

本文編號：1799984