Real-time 3D reconstruction method and system for large-scale scenes based on line-of-sight update algorithm
A real-time three-dimensional, large-scale technology, applied in radio wave measurement systems, calculations, 3D image processing, etc., can solve problems such as inability to achieve real-time three-dimensional reconstruction, poor accuracy and robustness, and insufficient use of each frame sensor. Achieve real-time high-quality reconstruction and improve quality
Active Publication Date: 2021-10-26
NANJING UNIV
View PDF5 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Image-based 3D reconstruction has poor accuracy and robustness in feature point matching, and cannot achieve real-time 3D reconstruction
[0004] At present, the reconstruction algorithm based on lidar mainly adopts the method of simplifying the point cloud and then triangulating it. It is difficult to achieve incremental update of the surface model, and it is impossible to achieve real-time processing of each frame of 3D point cloud data.
In addition, the reconstruction quality is low due to the underutilization of the pose information of the sensor in each frame
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0041] like figure 1 As shown, the present invention is based on a large-scale scene real-time three-dimensional reconstruction method based on the sight update algorithm, including the following steps:
[0042] Step 1: Sensors External Bragging Calculation.
[0043] The spatial absolute coordinates obtained by the current global positioning system are integrated as the initial value, and the acceleration obtained by the inertial navigation element (IMU) is integrated, and the current sensor is extracted, or the characteristic point extraction and matching of the radar point cloud per frame. Calculate the outer ginseng data of the laser radar at this time.
[0044] Step 2: The sight algorithm calculates the symbol distance value.
[0045] The space is performed in a certain range of space in the spatial position of the current laser radar. The body data structure is:
[0046] (p k，n , D k，n W k，n )
[0047] Where P k，n , 3D spatial coordinates of the nth body in the kth frame data s...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a large-scale scene real-time three-dimensional reconstruction method and system based on a line of sight update algorithm, belonging to the fields of computer vision and robotics. The problem solved by the present invention is: Aiming at the problem that the general algorithm cannot realize the real-time reconstruction of large-scale point cloud data of laser radar, a method of generating line of sight according to the current 3D point cloud, realizing data update, and thus real-time reconstruction is proposed. It mainly includes 3D point cloud acquisition and sensor external parameter data calculation, calculation of signed distance value of implicit surface based on line-of-sight algorithm, weighted fusion of signed distance value, volume rendering and storage of volume data, and real-time display of reconstruction effect. The present invention updates the implicit surface in real time by introducing a line-of-sight update algorithm, thereby realizing real-time reconstruction of large-scale scenes based on depth sensors such as radar, which not only has great advantages in speed, but also achieves better results in reconstruction quality.
Description
Technical field [0001] The invention belongs to the field of computer vision and robot, involving a three-dimensional reconstruction method and system based on large-scale scene real-scale scenes based on sight update algorithms. Background technique [0002] Real-time 3D reconstruction refers to the process of processing three-dimensional dot cloud or image data in real time, and generating a three-dimensional model. 3D reconstruction has a very important role in computer vision and the field of robots, is the basis for virtual reality, intelligent monitoring, robotic path planning. Since the laser radar can be collected from hundreds of thousand points per second, the amount of data is huge, and the data acquired by the data contains large noise, and the three-dimensional model of the large-scale scene is very challenging in real time. [0003] 3D reconstruction is mainly divided into two categories: 3D reconstruction based on camera images and 3D reconstruction based on laser ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06T15/08G01S17/894
CPCG01S17/89G06T15/08
Inventor 周余章坚于耀
Owner NANJING UNIV