CN-122023396-A - MBR aeration uniformity detection method and system based on image analysis

CN122023396ACN 122023396 ACN122023396 ACN 122023396ACN-122023396-A

Abstract

The invention belongs to the technical field of image data processing, and particularly relates to an MBR aeration uniformity detection method and system based on image analysis, wherein the method comprises the steps of constructing a Hessian matrix of an input image on the surface of an MBR pool, and determining the spot feature ratio and the structural strength of pixel points; determining the spot degree of the pixel points, determining the relative protrusion degree of the pixel points, determining the morphological background correction factor of the pixel points, determining the response threshold value after the pixel points are self-adaptive, obtaining a bubble binarization mask image, determining the bubble coverage rate of each subarea in the bubble binarization mask image, obtaining an aeration uniformity index, and judging whether aeration in an MBR pool is uniform. According to the invention, through analyzing morphological characteristics of pixel points and self-adapting to a response threshold value, the misjudgment limitation of the traditional Hessian matrix detection algorithm under wave interference is overcome, the air bubble segmentation is realized, and the accuracy and the robustness of the MBR pool aeration uniformity detection are improved.

Inventors

ZHANG XIAOLONG
ZHOU LIN
ZHANG MAI
GUO QING
JIANG BO

Assignees

陕西蔚蓝节能环境科技集团有限责任公司

Dates

Publication Date: 20260512
Application Date: 20260410

Claims (10)

1. An MBR aeration uniformity detection method based on image analysis is characterized by comprising the following steps: Constructing a Hessian matrix of an input image on the surface of the MBR pool, and determining the spot feature ratio and the structural strength of the pixel points according to two feature values of the Hessian matrix; determining the spot degree of the pixel point according to the spot feature ratio and the structural strength of the pixel point; determining the relative protrusion degree of the pixel point according to the difference value between the input image and the reconstructed image on the surface of the MBR pool; Determining morphological background correction factors of the pixel points according to the spot degree, the relative protrusion degree and the area variation coefficient of the connected domain to which the pixel points belong; correcting the response threshold value of the pixel point reference according to the morphological background correction factor to obtain a response threshold value after pixel point self-adaption; comparing a determinant response graph of the Hessian matrix with the self-adaptive response threshold to obtain a bubble binarization mask image, determining the bubble coverage rate of each subarea in the bubble binarization mask image, acquiring an aeration uniformity index according to the bubble coverage rate, and judging whether aeration in the MBR pool is uniform or not according to the aeration uniformity index.
2. The MBR aeration uniformity detection method based on image analysis according to claim 1, wherein the method for acquiring the input image of the MBR pool surface is characterized in that an RGB image of the MBR pool surface is converted into a gray image, the gray image is enhanced by using a CLAHE algorithm, and the obtained result is used as the input image of the MBR pool surface.
3. The MBR aeration uniformity detection method based on image analysis according to claim 1 is characterized in that the method for obtaining the speckle characteristic ratio and the structural strength is that two characteristic values of Hessian matrixes of all pixel points in an input image of the MBR pool surface are calculated, the ratio of the characteristic value with the smallest absolute value to the characteristic value with the largest absolute value is used as the speckle characteristic ratio of the pixel points, and the square root of the sum of squares of the two characteristic values of the Hessian matrixes of all the pixel points in the input image of the MBR pool surface is used as the structural strength of the pixel points.
4. A method for detecting the aeration uniformity of an MBR based on image analysis according to claim 1, 2 or 3, wherein the speckle degree satisfies the expression: ; In the formula, Is the first The degree of speckle of the individual pixels, Is the first The speckle characteristic ratio of the individual pixel points, Is the first The structural strength of the individual pixel points, As a function of the natural index of refraction, Is the maximum and minimum normalization function.
5. The MBR aeration uniformity detection method based on image analysis according to claim 1, wherein the relative protrusion degree acquisition method is characterized in that an input image of the surface of an MBR tank is used as a mask image, an image obtained by subtracting a preset height threshold from the input image of the surface of the MBR tank is used as a mark image, morphological gray level reconstruction is performed on the mask image by using the mark image to obtain a reconstructed image, and the difference value between the input image of the surface of the MBR tank and the reconstructed image is used as the relative protrusion degree of a pixel point.
6. The MBR aeration uniformity detection method based on image analysis according to claim 1 is characterized in that the area variation coefficient acquisition method comprises the steps of taking pixel points with relative salience larger than zero as foreground pixel points, index marking all the foreground pixel points by utilizing an eight-neighborhood connected component marking algorithm, taking a pixel point set with the same index mark as a connected domain to which the pixel points belong, calculating Euclidean distances from edge pixel points of the connected domain to the geometric center of the connected domain to which the pixel points belong, and taking the ratio of the standard deviation of Euclidean distances from all the edge pixel points to the geometric center of the connected domain to the average value of Euclidean distances from all the edge pixel points to the geometric center of the connected domain as the area variation coefficient of the pixel points.
7. The method for detecting the aeration uniformity of the MBR based on image analysis according to claim 1,5 or 6, wherein the morphological background correction factor satisfies the expression: ; In the formula, Is the first Morphological background correction factors for individual pixels, Is the first The degree of speckle of the individual pixels, Is the first The relative salience of the individual pixel points, Is the first The gray values of the individual pixels in the input image, Is the first The area variation coefficient of the connected domain to which each pixel point belongs, In order to prevent the denominator from being zero in the hyper-parameter, As a function of the natural index of refraction, Is the maximum and minimum normalization function.
8. The method for detecting the aeration uniformity of the MBR based on image analysis as recited in claim 1, wherein the adaptive response threshold satisfies the expression: ; In the formula, Is the first An adaptive response threshold for each pixel point, A global reference threshold preset for the Hessian matrix detection algorithm, Is the first Morphological background correction factors for individual pixels, Is the first And the adaptive response threshold value of each pixel point is used for adjusting the coefficient.
9. The method for detecting the aeration uniformity of the MBR based on image analysis of claim 1, wherein the determining whether the aeration in the MBR tank is uniform comprises, in response to the aeration uniformity index being less than a preset aeration uniformity pre-warning threshold, non-uniform aeration in the MBR tank.
10. An image analysis-based MBR aeration uniformity detection system comprising a processor and a memory, the memory storing computer program instructions that when executed by the processor implement an image analysis-based MBR aeration uniformity detection method according to any one of claims 1-9.

Description

MBR aeration uniformity detection method and system based on image analysis Technical Field The invention relates to the technical field of image data processing. More particularly, the invention relates to an MBR aeration uniformity detection method and system based on image analysis. Background Along with the increasing severity of water resource shortage and water environment pollution problems, MBR technology is gradually applied in the fields of sewage treatment and reuse due to the advantages of good water quality, small occupied area and the like, and in MBR process operation, the membrane flux maintenance of a membrane component mainly depends on the cross flow scouring of a gas-liquid two-phase flow generated by an aeration system on the surface of a membrane wire so as to slow down the formation of membrane pollution, so that the aeration uniformity is a key factor for guaranteeing the long-term stable operation of an MBR system, reducing energy consumption and prolonging the service life of the membrane. At present, monitoring of MBR aeration uniformity mainly depends on manual inspection, the manual inspection is low in efficiency and high in subjectivity, quantization cannot be achieved, in recent years, an image detection technology based on machine vision is gradually introduced into the field, and a traditional Hessian matrix detection algorithm is generally adopted for identifying bubbles in an existing image analysis method. In order to ensure that the high-concentration sludge mixed liquid can be fully fluidized and effectively scrub membrane wires, an MBR tank usually adopts a high-intensity aeration mode, the strong aeration working condition can cause a severe rolling phenomenon of the liquid level to form a large number of rough waves, ridge lines and tips of the waves can generate severe gray gradient changes in images, and high-frequency gradient noise generated by the waves is very similar to bubbles in shape. The conventional Hessian matrix detection algorithm mainly depends on the second derivative characteristic of pixel points, when the image is processed, the gradient characteristic of waves is easily misjudged as the edge of the bubbles, so that false overestimation of the number of the bubbles is caused, the real aeration bubbles and background wave noise cannot be accurately distinguished, the system is difficult to accurately identify local uneven aeration caused by blockage or breakage of an aeration pipe, and the operation optimization and fault early warning of an MBR system are influenced. Disclosure of Invention In order to solve the technical problems that the gradient noise generated by waves is easily misjudged as bubbles under the MBR Chi Jiang aeration working condition by the traditional Hessian matrix detection algorithm, so that false overestimation of bubble coverage rate is caused, and local aeration non-uniformity cannot be accurately identified, the invention provides a scheme in the following aspects. The invention provides an MBR aeration uniformity detection method based on image analysis, which comprises the steps of constructing a Hessian matrix of an input image on the surface of an MBR pool, determining the spot characteristic ratio and the structural strength of pixel points according to two characteristic values of the Hessian matrix, determining the spot degree of the pixel points according to the spot characteristic ratio and the structural strength of the pixel points, determining the relative protrusion degree of the pixel points according to the difference value between the input image and a reconstructed image on the surface of the MBR pool, determining a morphological background correction factor of the pixel points according to the spot degree and the relative protrusion degree of the pixel points and the area variation coefficient of a connected domain to which the pixel points belong, correcting the response threshold value of a pixel point reference according to the morphological background correction factor, obtaining a response threshold value after the pixel point self-adaption, comparing the line-column response graph of the Hessian matrix with the response threshold value after the self-adaption to obtain a bubble binarization image, determining the bubble coverage rate of each subarea in the bubble binarization image, obtaining the aeration uniformity index according to the difference value, and judging whether the aeration uniformity index is uniform or not in the aeration pool. The method highlights the spot-shaped structural characteristics of real bubbles by calculating the spot degree and utilizing the difference of spot characteristic ratio and structural strength, analyzes the local fluctuation and irregular shape of the wave background by calculating the relative protrusion degree and area variation coefficient, realizes the accurate identification of wave noise, improves the anti-interference capability of a Hessian matri