Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
WATER QUALITY IMAGE ANALYSIS METHOD AND SYSTEM BASED ON DEEP LEARNING, AND DEVICE AND MEDIUM
Document Type and Number:
WIPO Patent Application WO/2022/217589
Kind Code:
A1
Abstract:
Disclosed are a water quality image analysis method and system based on deep learning, and a device and a medium. A trained restricted Boltzmann machine model is used to acquire a dynamic weight value, and then grayscale processing is performed according to the weight value and a weighted average method, such that the adaptability of a dynamic image to an actual environment is improved, and the accuracy of dynamic image information is improved; then image abnormity determination is performed, such that the impact on the result of water quality analysis caused by the fact that other factors cause an image abnormity is eliminated, and the accuracy of water quality analysis is improved; noise detection and filtering processing are performed, such that the interference prevention capability of the image is greatly improved, the image analysis speed is improved, and the accuracy of water quality analysis is further improved; and normalization processing is performed on all pixel points of a normal image, such that the accuracy of dynamic image information is further improved, and the accuracy of water quality analysis is further improved.

Inventors:
YAN FENG (CN)
ZENG XIANGJI (CN)
PAN YAN (CN)
YANG JINGYA (CN)
TAN XIANGTIAN (CN)
Application Number:
PCT/CN2021/087813
Publication Date:
October 20, 2022
Filing Date:
April 16, 2021
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
CINF ENG CO LTD (CN)
International Classes:
G01N21/55; G06K9/00; G01N21/59; G06N3/04
Foreign References:
CN110687050A2020-01-14
CN112418128A2021-02-26
CN111047073A2020-04-21
CN108169445A2018-06-15
Attorney, Agent or Firm:
CHANGSHA ZONEKEY PATENT LAW FIRM (CN)
Download PDF: