Paper
14 April 2022 Urine sediment image recognition method based on re-parameterization network
Zhangwei Wu, Qingbo Ji
Author Affiliations +
Proceedings Volume 12178, International Conference on Signal Processing and Communication Technology (SPCT 2021); 121780B (2022) https://doi.org/10.1117/12.2631837
Event: International Conference on Signal Processing and Communication Technology (SPCT 2021), 2021, Tianjin, China
Abstract
Urine routine examination has moved towards automation in recent years. In order to improve the accuracy and recognition speed of the automatic urine analyzer, a urine sediment classification method based on the attention mechanism and reparameterization is proposed. Aiming at the problem of unclear images of some urine sediments and noise, image processing technology is used to pre-processed. Aiming at the problem of some classes are easy to be confused in classification, during network training, multiple branches are used for feature extraction, and the attention mechanism is used to focus on the key points in the image, which improves the accuracy of easily confused classes. Multi-branch can be equivalently replaced with a single branch during inference to improve the recognition speed. Experimental results show that the designed network can effectively improve the classification accuracy of urine sediment images.
© (2022) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Zhangwei Wu and Qingbo Ji "Urine sediment image recognition method based on re-parameterization network", Proc. SPIE 12178, International Conference on Signal Processing and Communication Technology (SPCT 2021), 121780B (14 April 2022); https://doi.org/10.1117/12.2631837
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Image classification

Image processing

Data modeling

Image enhancement

Blood

Network architectures

Chemical analysis

Back to Top