We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
基于语义分割的图像拼接实验设计.
- Authors
陈立伟; 宋涛羽
- Abstract
"Signal and information processing innovation practice" course is a comprehensive practice course of innovation and entrepreneurship. According to the technical requirements of the course, this paper designs an experimental case of image mosaic based on semantic segmentation. In order to improve the ability of stitching to distinguish foreground objects in image, the optimal stitching algorithm based on semantic segmentation is studied. CRF-RNN semantic segmentation is used to extract the foreground information of the image, and the extracted information is added to the energy matrix of YUV color space, so that the searched suture can bypass the foreground object. The simulation results show that the algorithm has a strong ability to distinguish the foreground objects in the image, and the mosaic results avoid the problems of compression and fragmentation, which is more in line with human visual perception. The design of the case can cultivate students' ability to apply the basic principles of professional knowledge, analyze complex engineering problems in the field of electronic information, and obtain effective conclusions.
- Publication
Experimental Technology & Management, 2021, Vol 38, Issue 12, p48
- ISSN
1002-4956
- Publication type
Article
- DOI
10.16791/j.cnki.sjg.2021.12.010