Liquid turbidity measuring device and method based on convolutional neural network

Abstract

The invention discloses a liquid turbidity measuring device and method based on a convolutional neural network. The measuring device comprises a backlight source fixing platform, a backlight source, a sample pool, an image acquisition device, an image acquisition device fixing support and a data processing unit, and the data processing unit further comprises a data storage unit, a turbidity judging unit and a display unit. According to the invention, the image acquisition device is used for acquiring characteristics and differences of the to-be-measured liquid image, the acquired liquid image data is processed through the convolutional neural network, a turbidity measurement result is given according to the characteristics and differences of the image, and simplification of data acquisition and rapidness of processing are realized by using the convolutional neural network; the device has the advantages of being simple in structure, low in cost, good in resolution ratio, intelligent and the like, and on-site measurement, rapid measurement and accurate judgment are easy to achieve.

Description

technical field [0001] The invention relates to the technical field of liquid turbidity analysis, in particular to a measuring device and method for measuring liquid turbidity based on a convolutional neural network. Background technique [0002] Since the beginning of the 20th century, the pace of industrialization in China has accelerated sharply, and the process of urbanization has also increased year by year, resulting in increasingly scarce water resources. According to statistics, the production of one ton of steel requires 5 tons or even more than ten tons of tap water. However, in reality, China is a country that suffers from drought and water shortage, and it is imminent to save water and protect water resources. Most of the large amount of wastewater produced by human beings in daily production and life is discharged into river waters after simple filtration treatment, especially untreated sewage, which seriously threatens the health of water bodies. On the one han...

AI Technical Summary

Problems solved by technology

[0003] Up to now, there are few equipment suitable for liquid turbidity analysis in special environments in the world, especially in some petrochemical and food industries, such as the paper industry, where the wastewater discharged presents a variety of colors, and the wastewater from the petroleum industry Contains a small amount of oil stains, etc. These factors, such as color and oil stains, will affect the precision and accuracy of traditional turbidity measurement devices, so they are not suitable for use in these industries. The feasible way is to purchase specific turbidity measurement devices, but these measurement devices Expensive and harsh conditions of use

In addition, most of the existing turbidity measurement equipment is expensive and has poor detection accuracy

[0004] In order to deal with the above-mentioned problems in measurement, a new detection method was proposed. For example, Chinese Invention Patent No. 201610620318.9 proposes a water quality image classification method, which uses algorithms to process the collected images to achieve the purpose of classification, but this method is limited to detection What kind of water pollution is this (algae bloom pollution, water hyacinth pollution, factory sewage pollution, domestic garbage pollution), and the specific turbidity value of the water body cannot be given, which limits the practicability of the system

Images