非平行双目视觉系统水下成像的极线约束模型

发布时间：2018-08-29 13:32

【摘要】：针对水下非平行双目成像系统图像匹配时无法进行平行校正,并且不再满足空气中的极线约束条件这一问题,推导出一种适用于水下的极线约束模型。根据摄像机和防水罩间相对位置关系的不同,把水下双目成像方式分为平行成像、共用折射面成像和独立折射面成像三种情况,并推导了后两种情况的极线约束模型。对摄像机进行标定,利用尺度不变特征变换(SIFT)进行立体匹配,推导了参考图上特征点在待匹配图上所对应的曲线,根据待匹配图上特征点是否在曲线上,剔除误匹配点,提高匹配精度。实验结果表明,利用所推导出的水下曲线约束模型,可以有效地剔除误匹配点,降低误匹配率。
[Abstract]:Aiming at the problem that the image matching of underwater non-parallel binocular imaging system can not be parallel corrected and can no longer satisfy the constraint condition of polar line in air, a polar line constraint model suitable for underwater is derived. According to the difference of the relative position relationship between the camera and the waterproof cover, the underwater binocular imaging mode is divided into three cases: parallel imaging, common refraction surface imaging and independent refraction surface imaging, and the polar line constraint model of the latter two cases is derived. The camera is calibrated and stereo matching is carried out by scale invariant feature transformation (SIFT). The curve corresponding to the feature points on the reference map is deduced. According to whether the feature points on the map are on the curve, the mismatched points are eliminated. Improve the matching accuracy. The experimental results show that the proposed model can effectively eliminate the mismatch points and reduce the mismatch rate.
【作者单位】：东北大学信息科学与工程学院;燕山大学工业计算机控制工程河北省重点实验室;
【基金】：河北省自然科学基金(D2015203310,D2014203153)
【分类号】：TP391.41

【相似文献】