Image stereo matching method
A stereo matching and image technology, applied in the field of image processing, can solve the problems of lack of texture, excessive reflection, difficult generalization ability, etc., and achieve the effect of great theoretical and practical value and wide application prospects.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment
[0065] (1) Build a training image library. There are certain difficulties in the acquisition of binocular images including depth, so there are relatively few binocular data sets currently in existence. At present, there are two main ways to obtain the depth map, one is to use lidar; the other is to use infrared depth sensor, the former obtains a sparse depth map, and the latter cannot work effectively outdoors. The lack of data limits the versatility of the algorithm. The current common practice is to pre-train on the synthetic data set first, and then fine-tune on a small number of real data sets. depth map, but the image itself lacks a certain realism. The lack of data greatly limits the generalization ability of the deep learning stereo matching algorithm, and often gets completely wrong results for scenes that have not been seen.
[0066] This part uses Secenflow, Kitti, Middlebury and other data in combination to provide sufficient data for model training. Furthermore,...
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