基于光场技术的三维测量系统的实现

发布时间：2018-01-12 04:11

本文关键词：基于光场技术的三维测量系统的实现　出处：《大连海事大学》2017年硕士论文　论文类型：学位论文

【摘要】：近年来,光场成像技术以其先拍照后聚焦、可变景深、可变视点等新颖成像特性受到学界的广泛关注。与传统相机的成像技术只能记录二维信息,缺乏对于物体深度信息的感知的缺点相比,光场成像技术由于其中的光场信息记录了光线的传播方向,蕴含着物体的三维信息,因而可以利用光场技术进行三维测量。本文将光场技术与主动式光源相结合,提出了一种新的光场三维测量方法,该方法克服了当前光场成像像素不足的缺点,大大降低了图像处理的复杂程度,实现了从图像信息采集系统的搭建到目标物体三维深度的重建。内容主要包括:(1)搭建图像数据采集系统,将光场成像系统与主动式光源相结合采集并记录不同深度的被测面板的光场信息;(2)对光场成像系统内参进行标定,获取光场成像系统模型参数;(3)对获取的光场数据进行处理,得到合成图片以及不同焦距的重聚焦图片;(4)对图像进行预处理,通过二值化,边缘检测,图像圆点的Hough变化提取主动光投射出的采样点的中心;(5)根据采样亮点中心所在的窗口像素的方差分布曲线求解出该亮点的像距,结合高斯成像公式求解深度,最后在三维坐标系中重建深度信息。实验结果证明了该方法的可行性和有效性。
[Abstract]:In recent years, the light field imaging technology to the first camera focus, depth of field variable, the variable viewpoint novel imaging characteristics has attracted wide attention in academic circles. And traditional camera imaging techniques can only record two-dimensional information, compared to the lack of depth information of the object perceived shortcomings of light field imaging technology because of the optical field propagation information recording the direction of light, contains the 3-D information of the object, so the 3D measurement can be performed using the light field technology. The light field technology and active light source combination, propose a new light field 3D measurement method, this method overcomes the deficiency of the current light field imaging pixel defects, greatly reduces the complexity of image processing, from the implementation of the image acquisition system is built to reconstruct the 3D depth of the target. The main contents include: (1) build the image data acquisition system, imaging system and light field The combination of active light source acquisition and measured panel light field information recorded in different depth; (2) the light field imaging system for internal calibration, obtain the light field imaging system parameters; (3) to deal with the light field data acquisition, synthetic images and different focal length focusing pictures (4;) preprocess the image by binarization, edge detection, sampling point Hough change extracting image dot active light projected in the center; (5) according to the variance of the distribution curve to solve sampling window where the center pixel highlights the bright spot like distance, combined with Gauss formula to solve the imaging depth, the final reconstruction depth the information in the 3D coordinate system. The experimental results show that the method is feasible and effective.

【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【相似文献】