Research on RGB-D Image Recognition Algorithms Based on Parallel Deep Learning

Authors

  • Xiaohui Zhang , Peng Xu . Tao Xu

Abstract

Deep learning has achieved remarkable results in image recognition, using RGB-D
image instead of simple feature connection. It is of great value to use RGB in depth
recognition. In addition, effective feature fusion can make better use of RGB-D image
instead of simple feature connection. It is of great value to use RGB in depth
recognition. Based on this, this paper first describes the concept of RGB-D image
recognition based on deep learning, and then analyzes the problems of the RGB-D
image recognition method based on deep learning, and studies the RGB-D image
recognition algorithm based on multi-mode hybrid structure, including RGB-D image
recognition frame, algorithm principle, features, feature extraction and preprocessing.
Finally, the RGB-D image recognition algorithm based on multi-mode hybrid structure
is studied, and the experiment of image recognition algorithm is given and verified by
experiment.

Downloads

Published

2020-11-01

Issue

Section

Articles