光学学报, 2018, 38 (1): 0112005, 网络出版: 2018-08-31
基于光条信度评价的线结构光传感器曝光时间优化 下载: 1095次
Exposure Time Optimization for Line Structured Light Sensor Based on Light Stripe Reliability Evaluation
测量 线结构光传感器 结构光条 高斯信度评价模型 曝光时间 measurement line structured light sensor structured light stripe Gaussian reliability evaluation model exposure time
摘要
研究了基于线结构光光条信度评价的相机曝光时间优化方法,将信度评价结果作为曝光时间优化的参考指标。结合光条切面服从高斯分布的特点,对已有光条信度评价模型进行改进,构建高斯信度评价模型;利用亚像素光条中心提取方法对不同曝光时间下采集到的光条图像进行光心提取,提取结果经高斯信度评价模型评价得到评价信度值C 和灰度值R ;分析曝光时间t 对评价结果的影响机理,构建t -C 、t -R 变化模型,并通过分析变化模型得到了相机最优曝光时间。实验以230 lx光强下的高精密齿条和水泥路面模型作为测量对象,结果表明:曝光时间优化后的测量更准确且效率更高,单光条测量的平均残差平方和约为0.03 mm
2,重构模型的平均构造深度与真实模型仅差9.8%。
Abstract
Optimization methods of camera's exposure time based on line structured light stripe reliability evaluation are studied, and the reliability evaluation results are used as reference indexes for optimization of exposure time. Firstly, existing light stripe reliability evaluation model is improved by combination with the characteristics that light stripe cross section obey Gaussian distribution, and then Gaussian reliability evaluation model is constructed. With the extraction method of sub-pixel light stripe center, light stripe centers are extracted from stripe images acquired with different exposure times. Extraction results are evaluated by Gaussian reliability evaluation model to obtain the values of evaluation reliability C and gray level R. Then, the influence mechanism of exposure time t on the evaluation results are analyzed respectively. Variation models of t-C and t-R are built, and the camera's optimal exposure time is finally obtained through analysis of variation models. Experiments take the high-precision spline and cement pavement model under light intensity of 230 lux as an experimental measurement object. Results show that measurement with the optimized exposure time is more reliable and more efficient. Mean residual sum of squares of the single light strip measurement is about 0.03 mm
2. The average texture depth of reconstructed model is only 9.8% away from actual model.
李涛涛, 杨峰, 李策, 方亮. 基于光条信度评价的线结构光传感器曝光时间优化[J]. 光学学报, 2018, 38(1): 0112005. Taotao Li, Feng Yang, Ce Li, Liang Fang. Exposure Time Optimization for Line Structured Light Sensor Based on Light Stripe Reliability Evaluation[J]. Acta Optica Sinica, 2018, 38(1): 0112005.