An End-to-End Point Cloud Registration Method Based on Feature Learning
A feature learning and point cloud registration technology, applied in neural learning methods, image analysis, image enhancement and other directions, can solve the problem of not being able to directly generate point cloud pair transformation matrix, increasing algorithm complexity, etc., to avoid falling into local optimum , reduce the probability of falling into the local optimal solution, improve the accuracy and efficiency
Active Publication Date: 2022-05-27
浙江大学计算机创新技术研究院
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, these methods are mainly used to find the correspondence between point clouds, and cannot directly generate the transformation matrix between point cloud pairs, which increases the complexity of the algorithm
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
[0036] The present invention will be further described below in conjunction with the accompanying drawings and specific implementations.
[0037] It should be understood that the embodiments described in the present invention are exemplary, and the specific parameters used in the description of the embodiments are only for the convenience of describing the present invention, and are not intended to limit the present invention.
[0038] like Figure 4 As shown, an embodiment of the present invention and its implementation process are as follows:
[0039] Step 1: Collect the point cloud of the object to be tested, each point of the point cloud is a coding point, and a spherical area is constructed in the point cloud with each coding point as the center; A point is called a code point. The neighborhood points of each code point are searched using the Ball Query method. The radius of the spherical region constructed by the spherical query is set to 0.3.
[0040] Step 2: Random...
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 end-to-end point cloud registration method based on feature learning. Use the neighborhood points of each point in the point cloud to construct the local geometric features of the point, and use the spatial coordinates, normal information and local geometric features of each point to construct the mixed features of the point; establish a template to process the point cloud and the source at the same time Point cloud end-to-end point cloud registration deep learning network; design translation loss function and rotation loss function, and complete the training and learning of point cloud registration network under the joint supervision of the two loss functions. The end-to-end point cloud registration method based on feature learning proposed by the present invention is not sensitive to the initial position of the rigid body transformation, reduces the probability of the algorithm falling into a local optimal solution, and can effectively improve the accuracy and efficiency of point cloud registration.
Description
technical field [0001] The invention relates to a physical point cloud processing method in the fields of computer artificial intelligence and three-dimensional point cloud registration, in particular to an end-to-end point cloud registration method based on feature learning. Background technique [0002] The point cloud registration task is mainly to find the rigid body transformation between two unknown corresponding point clouds, which is widely used in reverse engineering, dimension measurement, robotics and other fields. The disorder of point clouds and the complex initial correspondence between different point clouds increase the difficulty of point cloud registration. The Iterative Closest Point (ICP) algorithm and its variants are widely used effective point cloud registration methods, but this method is very sensitive to the initial corresponding position of the point cloud pair, and it is easy to fall into local optimum. In addition, the continuous iteration metho...
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): G06T7/33G06N3/04G06N3/08
CPCG06T7/33G06N3/08G06T2207/10028G06T2207/20081G06T2207/20084G06N3/045
Inventor 宋亚楠沈卫明陈刚
Owner 浙江大学计算机创新技术研究院