激光与光电子学进展, 2023, 60 (1): 0112002, 网络出版: 2022-12-09
运动模糊情况下的结构光光条中心快速提取 下载: 626次
Fast Extraction of Structure Light Strip Center Under Motion Blur
测量与计量 线结构光 运动模糊 中心提取 阈值分割 measurement and metrology linear structured light motion blur center extraction threshold segmentation
摘要
为了在复杂的运动模糊情况下快速提取光条中心,提出了一种新的结构光光条中心快速提取算法。通过分析线结构光运动成像模糊的原因和图像中光条截面的灰度值分布规律,设计了基于理论光条成像宽度的P-tile阈值分割算法,以解决运动模糊情况下光条成像宽度不一、亮度无规则变化引起的光条区域提取困难问题。根据光条图像的特点,通过改进的区域生长算法提高光条定位速度,根据光条截面的灰度值分布特点提取结构光光条中心。实验结果表明,在运动模糊情况下,改进的P-tile阈值分割算法具有精度高、速度快的优点,相比极值法、大津法能更好地分割出有效光条区域,在结构光工业高速测量领域具有实用价值。
Abstract
In this paper, a new fast center extraction algorithm of structured light stripe is proposed, which can extract the center of structured light stripe quickly under complex motion blur. This paper analyzes the causes of blurred motion imaging of line structured light and the gray value distribution law of the light stripe section in the image, and designs a P-tile threshold segmentation algorithm based on the theoretical light stripe imaging width, which solves the difficulty of extracting the light stripe area caused by the different light stripe imaging width and irregular brightness change in the case of motion blur. According to the characteristics of the light stripe image, the positioning speed of the light stripe is improved by the improved region growth algorithm, and the center of the structured light stripe is extracted according to the gray value distribution of the light stripe section. The experimental results show that the improved P-tile threshold segmentation algorithm is more effective than the extreme value method and Otsu method when segmenting the optical stripe region under motion blur. Our center extraction method provides faster speed and better precision, and has practical value in high-speed industrial measurement of structured light.
张倩, 张坤, 朱美强, 李海港, 王军. 运动模糊情况下的结构光光条中心快速提取[J]. 激光与光电子学进展, 2023, 60(1): 0112002. Qian Zhang, Kun Zhang, Meiqiang Zhu, Haigang Li, Jun Wang. Fast Extraction of Structure Light Strip Center Under Motion Blur[J]. Laser & Optoelectronics Progress, 2023, 60(1): 0112002.