测绘与空间地理信息
測繪與空間地理信息
측회여공간지리신식
GEOMATICS & SPATIAL INFORMATION TECHNOLOGY
2015年
6期
69-71
,共3页
关键帧%图像拼接%特征匹配%光束法平差
關鍵幀%圖像拼接%特徵匹配%光束法平差
관건정%도상병접%특정필배%광속법평차
key frames%image stitching%feature matching%bundle adjustment
以关键帧作为待拼接图像，为了减少图像间的扭曲和变形，保证图像整体关系的正确性，减少误差的累积传播，得到最佳的拼接结果，本文采用三种技术来提高图像自动拼接的精度。首先，通过摄像机标定实现关键帧的矫正；其次，采用基于Harris和SURF算法完成特征点的匹配；最后，在特征匹配的基础上进行光束法平差得到高精度的图像参数。实验结果表明，此方法稳定可靠，拼接精度高。
以關鍵幀作為待拼接圖像，為瞭減少圖像間的扭麯和變形，保證圖像整體關繫的正確性，減少誤差的纍積傳播，得到最佳的拼接結果，本文採用三種技術來提高圖像自動拼接的精度。首先，通過攝像機標定實現關鍵幀的矯正；其次，採用基于Harris和SURF算法完成特徵點的匹配；最後，在特徵匹配的基礎上進行光束法平差得到高精度的圖像參數。實驗結果錶明，此方法穩定可靠，拼接精度高。
이관건정작위대병접도상，위료감소도상간적뉴곡화변형，보증도상정체관계적정학성，감소오차적루적전파，득도최가적병접결과，본문채용삼충기술래제고도상자동병접적정도。수선，통과섭상궤표정실현관건정적교정；기차，채용기우Harris화SURF산법완성특정점적필배；최후，재특정필배적기출상진행광속법평차득도고정도적도상삼수。실험결과표명，차방법은정가고，병접정도고。
In order to reduce the distortion and deformation between the images ,with key frames as for stitching images ,ensure the cor-rectness of the overall relationship between images , reduce the accumulation of error propagation , and ensure that the best stitching re-sults, this paper takes three kinds of technology to improve the accuracy of image mosaic automatically .Firstly, achieving the key frames by the camera calibration correction .Secondly , using Harris and SURF to complete the matching based on feature points .Fi-nally, getting high precision parameters on the basis of the feature matching method of bundle adjustment .The results show that the al-gorithm in this paper is stable and reliable , and have high precision of mosaic .
