【作者】 常春；

【导师】 徐林；

【作者基本信息】 东北大学 ， 控制理论与控制工程， 2019， 硕士

【摘要】 根据获取的二维图像信息来对物体进行三维重构具有非常高的应用价值。从宏观上看,它可以用来对浩瀚的宇宙进行探索,从微观上看,它可以应用在对分子、原子的研究。三维重构技术已经广泛地应用到了日常生活中,并且正在改变着人类的生活质量以及方式,如医学诊断,生产装配,机器人视觉等方面,给人类的生产和生活带来很大的便利。由于高校和研究所的研究环境提升,设备与人数增多,产生了实验室人员、设备的监控管理和实验室安全监控的需求。本研究通过三维重建技术实现三维可视化在线监控管理实验室,提升实验室对设备和人员的管理效率,保障实验室安全。本研究使用基于改进DIBR算法的三维重构技术作为基础,构建实验室三维可视化监控管理平台。DIBR算法(Depth Image Based Rendering)是基于深度图的图像绘制算法,该算法具有绘制速度快,并且无需训练、重构结果分辨率高等优点。本研究根据DIBR的技术特点,提出一种改进的DIBR算法,通过双视点逆向映射能够有效解决重构的图像中出现的空洞、伪轮廓等问题,提高算法的精度。研究首先对图像进行降噪滤波处理,降低噪声对三维重建产生的干扰,将处理的图像数据作为DIBR算法的输入数据,然后使用摄像机参数标定得到内参数矩阵和多角度目标图像,通过DIBR算法,计算出目标对象不同角度的图像数据,为三维重建提供图像数据,并能减少占用数据存储空间,为实验室管理监控平台提高效率。最后利用准备好的图像数据进行SFM算法三维重建,通过SIFT特征提取算法去提取图像特征,匹配得到初始匹配对后,再利用RANSAC算法剔除误匹配得到图像的匹配点,计算出各个特征点的三维坐标,生成三维点云,最终重构出三维模型。基于改进DIBR算法的三维重构技术应用于实验室三维可视化能够实现对人员和设备位置的监控,实现实验室三维可视化监控管理,提高实验室管理效率。更多还原

【Abstract】 The 3D reconstruction technology based on the two-dimensional image information has a very high application value.From a macro perspective,it can be used to explore the vast universe.From the microscopic point of view,it can be applied to the study of molecules and atoms.The 3D reconstruction technology has been widely used in daily life,and is changing the quality of life and methods of human beings,such as medical diagnosis,production assembly,robot vision,etc.,which bringing great convenience to human production and life.Due to the improved research environment of universities and research institutes,the increase in equipment、the number of people has created the need for monitoring and management of laboratory personnel and equipment and laboratory safety monitoring.This research realizes the 3D visualization online monitoring and management laboratory through 3D reconstruction technology,improves the management efficiency of the equipment and personnel in the laboratory,and guarantees the safety of the laboratory.This study uses the improved DIBR algorithm based 3D reconstruction technology as the basis to build a 3D visualization monitoring management platform.The DIBR algorithm(Depth Image Based Rendering)is an image rendering algorithm based on depth map,which has the advantages of fast drawing speed,no need for training,and high resolution of reconstruction results.According to the technical characteristics of DIBR,this paper proposes an improved DIBR algorithm.The two-view reverse mapping can effectively solve the problems of voids and false contours in reconstructed images and improve the accuracy of the algorithm.The research firstly performs noise reduction filtering on the image to reduce the interference of noise on the 3D reconstruction,and uses the processed image data as the input data of the DIBR algorithm.In the second step,the internal parameter matrix and the multi-angle target image are obtained by using the camera parameter calibration,and the image data of the target object at different angles is calculated by the DIBR algorithm,and the image data is provided for the three-dimensional reconstruction,and the occupied data storage space can be reduced.Manage monitoring platforms to increase efficiency.In the third step,the SFM algorithm 3D reconstruction is performed by using the prepared image data,and the image features are extracted by the SIFT feature extraction algorithm.After matching the initial matching pairs,the RANSAC algorithm is used to eliminate the matching points of the mismatched images and calculate the feature points.The three-dimensional coordinates,generate a three-dimensional point cloud,and finally reconstruct a three-dimensional model.The 3D reconstruction technology based on the improved DIBR algorithm can be applied to the 3D visualization of the laboratory to monitor the position of personnel and equipment,it realize the 3D visual monitoring and management of the laboratory,and improve the efficiency of laboratory management.更多还原