RGB-DSLAM method and device based on key frame
A key frame and current frame technology, applied in image data processing, instrumentation, calculation, etc., can solve problems such as large errors and large memory, and achieve the effect of solving SLAM problems, reducing the number of stored frames, and reducing mileage measurement errors
Inactive Publication Date: 2020-02-18
BEIJING YINGPU TECH CO LTD
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] However, in the above-mentioned prior art, since the storage method mostly adopts the form of point cloud, a large amount of memory is required, and due to the use of key frames, the error is often large
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
[0040] figure 1 is a flow chart of a key frame-based RGB-D SLAM method according to an embodiment of the present application. see figure 1 , the method includes:
[0041] 101: Obtain an RGB-D color and depth image from a dataset;
[0042] 102: Use FOVIS to perform pose estimation based on color and depth images, obtain a relative transformation matrix between two poses, and create a point cloud image according to the relative transformation matrix;
[0043] 103: Use the g2o framework in the process of closed-loop detection to select key frames, create nodes corresponding to key frames in the point cloud graph, store nodes and perform optimized calculations.
[0044] In this embodiment, optionally, create a point cloud image according to the relative transformation matrix, including:
[0045] According to the relative transformation matrix, use Eqs and use mesh filter to create point cloud map.
[0046] In this embodiment, optionally, key frames are selected, including:
...
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 an RGB-DSLAM method and device based on a key frame, and relates to the field of SLAM. The method comprises the following steps: acquiring RGB-D color and depth images from a data set; performing attitude estimation by using FOVIS according to the color and depth images to obtain a relative transformation matrix between two poses, and creating a point cloud graph accordingto the relative transformation matrix; and in a closed-loop detection process, selecting a key frame by utilizing a g2o framework, creating a node corresponding to the key frame in the point cloud graph, storing the node, and carrying out optimization calculation. The device comprises an acquisition module, a creation module and a detection module. According to the method, an existing framework isoptimized, the number of necessary storage frames is reduced, high robustness is achieved, the SLAM problem can be effectively solved, mileage measurement errors are reduced, and a coherent three-dimensional map is generated.
Description
technical field [0001] The present application relates to the field of SLAM, in particular to a key frame-based RGB-DSLAM method and device. Background technique [0002] RGB-D (color and depth imagery) cameras for robotics research are relatively new. Henry et al. developed a RGB-D SLAM (Simultaneous Localization And Mapping, instant positioning and map construction) system in 2012, combining sparse visual features and ICP (Iterative Closest Point, Iterative Closest Point), using the concept of key frames to construct and optimize a location map. In 2011, Huang et al. implemented a visual odometry system called FOVIS (Fast Odom-etry from VISion) on a micro-aircraft, which has an RGB-D camera and uses sparse visual features for positioning and drawing. In 2014, Silva and Goncalves developed a visual odometry and mapping system for RGB-D cameras that only required the use of a CPU. Another RGB-D SLAM system was developed by Endres et al. in 2012, using sparse visual featu...
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 IPC IPC(8): G06T7/90G06T7/593G06T7/70
CPCG06T7/90G06T7/70G06T7/593G06T2207/10028
Inventor 吉长江
Owner BEIJING YINGPU TECH CO LTD