| 基于点云特征提取与匹配算法的智能电表检定线机器人位姿校正方法 |
 点此下载全文 |
| 引用本文：鲁观娜，李亮，王慧楠，吕言国，高帅.基于点云特征提取与匹配算法的智能电表检定线机器人位姿校正方法[J].计算技术与自动化,2025,(3):83-87 |
| 摘要点击次数: 284 |
| 全文下载次数: 214 |
|
|
| 中文摘要:机器人残余位姿差补偿过程忽略了图像缩放对特征提取映射关系的影响，导致校正精度不佳。为此，提出了基于点云特征提取与匹配算法的智能电表检定线机器人位姿校正方法。结合FAST关键点算法，对圈定范围内的点云特征进行提取，从而规避尺度恒定性问题，并构建出正态分布地图；对特征点误匹配的概率进行校正，求解机器人实际位姿，通过对位姿特征点与正态分布地图之间的相关性，将位姿增量补偿到实际位姿信息中，实现位姿校正。实验结果表明，在复杂环境和动态变化条件下，位姿增量均较低，具有较高的准确性和稳定性。对机器人位姿进行校正时，具备较高的校正精度。 |
| 中文关键词:点云特征 特征匹配 机器人 位姿校正 |
| |
| Smart Meter Checking Line Robot Position Correction Method Based on Point Cloud Feature Extraction and Matching Algorithm |
|
|
| Abstract:The residual pose difference compensation process of robots ignores the impact of image scaling on feature extraction mapping relationships, resulting in poor correction accuracy. A robot pose correction method for intelligent meter calibration line based on point cloud feature extraction and matching algorithm is proposed. Combining the FAST key point algorithm, extract point cloud features within the designated range to avoid the problem of scale invariance and construct a normal distribution map; Correct the probability of mismatched feature points, solve the actual pose of the robot, and compensate the pose increment to the actual pose information by comparing the correlation between the pose feature points and the normal distribution map, achieving pose correction. The experimental results show that under complex environments and dynamic changing conditions, the pose increment is low, and it has high accuracy and stability. When correcting the robot’s posture, it has a high correction accuracy. |
| keywords:point cloud features feature matching robot positional correction |
| 查看全文 查看/发表评论 下载pdf阅读器 |
位访问者,今日访问
