论文部分内容阅读
提出一种考虑特征分布的局部特征提取算法,提取非结构化环境中符合各项同性分布的局部特征作为自然路标,使得基于行为的机器人能够利用这些路标实现高精度的视觉归航.以SIFT(scale invariant feature transform)算法为基础,通过改善局部特征的分布均匀性得到了UD-SIFT(uniform distribution-SIFT)算法,并提出一种新的评价标准对局部特征的分布均匀性进行评估.采用基于全景视觉的ADV(average displacement vector)和ALV(average landmark vector)方法,在室内、走廊和室外环境下进行归航实验,相对于原始SIFT算法,采用UD-SIFT算法时,归航的平均方向误差降低了25.01%以上.结果表明:本文算法有效改善了特征的均匀分布状况,提高了机器人的归航精度.
A local feature extraction algorithm considering the feature distribution is proposed to extract the local features that conform to the same-sex distribution in the unstructured environment as natural landmark, so that the behavior-based robot can use these landmarks to achieve high-precision visual navigation. scale invariant feature transform algorithm, the uniform distribution-SIFT (UD-SIFT) algorithm is obtained by improving the distribution uniformity of local features and a new evaluation criterion is proposed to evaluate the distribution uniformity of local features. Compared with the original SIFT algorithm and the UD-SIFT algorithm, the average directional error of the homing is compared with the average displacement vector (ADV) and the average landmark vector (ALV) method of panoramic vision. Which is reduced by more than 25.01% .The results show that this algorithm can effectively improve the uniform distribution of features and improve the homing accuracy of the robot.