Binocular vision three-dimensional scanning system
A three-dimensional scanning and binocular vision technology, applied in the field of three-dimensional scanning, can solve cumbersome problems, and achieve the effects of easy portability, improved use distance and range, and rapid molding.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0028] A binocular vision three-dimensional scanning system, please refer to the appendix figure 1 -Attached figure 2 As shown, it includes a scanner and a data processing center 9. The scanner includes a first casing 1 and a second casing 2. The second casing 2 is provided at the rear end of the first casing 1 and is connected with the first casing 1. The housing is formed in one piece, the first housing 1 is provided with a binocular camera 3, and the second housing 2 is provided with an analog-to-digital conversion module 4, a data processing module 5, a data storage module 6, a wireless transmission module 7 and a microprocessor 8, the binocular camera 3 is electrically connected with the analog-to-digital conversion module 4, the data processing module 5 is electrically connected with the analog-to-digital conversion module 4 and the data storage module 6 respectively, and the microprocessor 8 is respectively connected with the wireless transmission module 7 is electric...
Embodiment 2
[0039] The data processing center 9 performs feature extraction (FEM), disparity initialization (DIM), and disparity refinement (DIM) on the received data, wherein FEM provides robust multi-scale features for DIM and DRM, and DIM uses in low-resolution levels A 3D convolutional neural network initializes disparity, and DRM uses a multi-branch fusion (MBF) module at multiple scales to progressively recover disparity map details.
[0040] Further, FEM adopts L group convolution to reduce the spatial resolution, and its structure is as follows image 3 As shown, the L groups of convolutions are denoted as FEM-l (l=1, 2, ..., L), which both contain two convolution layers with a kernel size of 3, and the convolution strides are 2 and 1, respectively. Represents image features output by FEM at multiple resolution levels. In order to ensure the operating efficiency of the algorithm and fuse multi-scale features at the same time, the SPP (spatial pyramid pooling) layer is introduced ...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com