Non-reference stereo image quality evaluation method based on local-to-global feature regression
A local regression, stereo image technology, applied in the field of image processing, can solve problems such as binocular competition and binocular masking
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0021] The present invention first assigns different labels to the image blocks of the left and right viewpoints through the feature similarity (FSIM) algorithm, uses the calculated labels to guide the local regression neural network of the left and right channels to perform pre-training at the same time, and uses the trained local regression neural network The parameters are saved. Then, the local regression neural network of the left and right channels is fused, and on the basis of the pre-training model, the subjective evaluation value (DMOS) is used as the label to guide the global regression neural network to train, and the parameters of the global regression neural network are fine-tuned. Implement global regression of features. The quality of the stereoscopic image is extracted and predicted by the trained global regression neural network.
[0022] Local RNN label generation:
[0023] The content of this work is partly based on the algorithm proposed in literature [9]...
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