Flocculating settlement observation system and method based on machine vision
By using a machine vision-based flocculation and sedimentation observation system and method, and leveraging industrial cameras and image processing technology, the problem of missing observation of flocculation and sedimentation effects has been solved, enabling precise control of flocculant dosage and reduction of reagent costs.
Patent Information
- Application Number
- CN202511112993.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-10
- Publication Date
- 2025-11-18
AI Technical Summary
The lack of effective methods to observe the flocculation and sedimentation effect makes it difficult to control the dosage of flocculants, increasing the cost of the agents.
A machine vision-based flocculation sedimentation observation system, including an industrial camera, a transparent observation chamber, and a strip light source, was adopted. The interface gradient value was determined through image processing and grayscale analysis, and the sedimentation velocity was evaluated by fitting a linear equation curve.
It enables low-cost and effective observation of flocculation and sedimentation effects, simplifies the control of flocculant dosage, and reduces reagent costs.
Smart Images
Figure CN120971286A_ABST
Abstract
Description
Technical Field
[0001] This patent application belongs to the field of flocculation and sedimentation observation technology, and more specifically, it relates to a flocculation and sedimentation observation system and method based on machine vision. Background Technology
[0002] With increasing energy demand, coal slurry from coal washing has become a significant environmental problem. The solid particles in coal slurry are difficult to settle, affecting water quality and environmental safety. To address this issue, researchers have studied the flocculation and settling characteristics of difficult-to-settle coal slurry. Flocculants can improve the settling properties of particles in coal slurry. Furthermore, appropriate flocculant dosage and extended settling time are also key factors in improving settling efficiency. These research findings provide important theoretical and practical guidance for the treatment of difficult-to-settle coal slurry.
[0003] Currently, flocculation and sedimentation are achieved by adding flocculants. However, there is a lack of a necessary method to observe the sedimentation effect in order to determine whether the flocculant dosage is reasonable and to avoid increasing the cost of the reagents due to exceeding the reasonable value. That is, we need to observe to judge whether the flocculation and sedimentation meet the requirements, while controlling the cost of the reagents at a low level. Summary of the Invention
[0004] The technical problem to be solved by the present invention is to provide a flocculation and sedimentation observation system and method based on machine vision, which solves the problem of missing feedback link of flocculation and sedimentation effect in related technologies.
[0005] To solve the above problems, the technical solution adopted by the present invention is as follows: A machine vision-based flocculation sedimentation observation system includes an industrial camera, a transparent observation chamber corresponding to the industrial camera, and a strip light source set next to the transparent observation chamber. The industrial camera is set vertically facing the front of the transparent observation chamber, and the transparent observation chamber contains a flocculation solution.
[0006] Furthermore, the industrial camera is a webcam, and the strip light source is a planar light source.
[0007] Furthermore, the transparent observation chamber is a cubic box made of acrylic material.
[0008] A machine vision-based method for observing flocculation and sedimentation, utilizing the aforementioned observation system, includes the following steps: Step 1: Use an industrial camera to take a vertical shot from the side of the transparent observation chamber to capture an image of the chamber. Step 2: Pre-define the region of interest in the image for analysis and calculation; Step 3: Perform grayscale processing on the region of interest in the image; Step 4: Within the region of interest of the image, calculate the average grayscale value of each row of pixels along the column direction (i.e., the Y direction) and place it in a list; Step 5: Calculate the gradient value of all grayscale average values in the y-direction from Step 4; Step 6: Determine the gradient value of the interface: Take the row with the smallest gradient value to represent the interface of settlement, record the height mark line and Y coordinate value of the interface. The Y coordinate value represents the height value of the interface. The gradient value can be positive or negative, and the smallest negative value represents the interface. Step 7: Repeat steps 1 to 6 to continuously capture N images of the transparent observation chamber, record the height values of N interfaces, and then proceed to step 8; Step 8: Fit two linear equation curves (y = a1*x+b1, y = a2*x+b2) using N height values to represent the trajectory of the interface descent, where the x-axis is the time interval, the y-axis is the position of the interface, and a1, b1, a2, and b2 are all coefficients. Step 9: Take the absolute value of the coefficient a1 in the first segment of the linear equation curve to represent the velocity index value of the interface descent. The velocity index value is used to compare the results of the settlement detection at different times.
[0009] Furthermore, in step 2, the region of interest in the image is selected as a rectangular area near the edge of the supplementary lighting side of the bar light source to improve the calculation accuracy.
[0010] Furthermore, in step 3, grayscale processing is performed using the average value, weighted average method, or maximum value method.
[0011] Furthermore, in step 7, the time interval for acquiring images of the transparent observation chamber is fixed, and is performed at a frequency of once per second.
[0012] Due to the adoption of the above technical solution, the beneficial effects achieved by this invention are: Using industrial cameras to collect data on flocculation and sedimentation in the observation chamber is a low-cost method that meets production requirements, is simple in design, and provides good observation results.
[0013] This invention also provides a matching observation system, which is easy to obtain materials for, simple to design and construct, and highly practical. The designed observation method is simple to implement, easy to observe, and produces good observation results, making it extremely valuable for practical use. Attached Figure Description
[0014] Figure 1 This is a flowchart of the visual inspection method of the present invention; Figure 2 This is a diagram showing the positions of the camera and the transparent observation chamber in this invention; Figure 3 This is a diagram showing the processing results of the images acquired in this invention; Figure 4 This is a fitted line graph of the present invention; Figure 5 This is a diagram of the fitting process of the present invention.
[0015] In the picture: 1. Industrial camera; 2. Transparent observation chamber; 3. Strip light source. Detailed Implementation
[0016] The present invention will be further described in detail below with reference to the embodiments.
[0017] A machine vision-based flocculation sedimentation observation system, such as Figure 2 As shown, it includes an industrial camera 1, a transparent observation chamber 2 corresponding to the industrial camera 1, and a strip light source 3 set next to the transparent observation chamber 2. The industrial camera 1 is set vertically facing the front of the transparent observation chamber 2, and the transparent observation chamber 2 contains a flocculation solution.
[0018] In terms of selection, industrial camera 1 is a webcam, and strip light source 3 is a planar light source. Transparent observation chamber 2 is a cubic box made of acrylic material.
[0019] A machine vision-based method for observing flocculation and sedimentation utilizes Figure 1 The system as a whole is like Figure 1 As shown, it includes the following steps: Step 1: As Figure 2 As shown, industrial camera 1 faces the front of transparent observation room 2 and takes a vertical picture to capture an image of transparent observation room 2; Step 2: As Figure 2 As shown, in order to enhance the recognizability of the boundary layer, strip light sources 3 are arranged on the adjacent surfaces of the imaging surface of the transparent observation chamber 2 in step 1; Step 3: As Figure 3 As shown in the rectangle in the image, the rectangular area on the edge of the transparent observation room 2 near the supplementary light source 3 is selected in the acquired image. This area is used to analyze and calculate the region of interest in the image. The region of interest image is then processed in grayscale, specifically by using the average value, weighted average method, or maximum value method. Step 4: Within the region of interest, calculate the average grayscale value of each row of pixels along the column direction (i.e., the y-direction) and place it in a list; Step 5: Calculate the gradient value of all grayscale average values from Step 4 in the column direction (y-direction); Step 6: The interface transitions from light to dark from top to bottom. This is reflected in the gradient values. The row with the smallest gradient value represents the settlement interface. The sequence number of the smallest value represents the location of that value, thus determining the gradient value of the interface: The row with the smallest gradient value represents the settlement interface. Record the height marker line and Y coordinate value of the interface. The Y coordinate value represents the height of the interface. Gradient values can be positive or negative, with the smallest negative value representing the interface. Step 7: As Figure 3 As shown, repeat steps 1 to 6. The interface position will be calculated in N consecutively acquired images of transparent observation room 2. The position is marked on the figure with a red line (or other colored line) and the position information is marked on the figure. After tracking and capturing N images and recording a total of N interface height values, proceed to step 8. The time interval for acquiring images of transparent observation room 2 is fixed and is performed at a frequency of 1 acquisition per second. Step 8: Fit two linear equation curves (y = a1*x+b1, y = a2*x+b2) using N height values to represent the process of the interface descending, where the x-axis is the time interval, the y-axis is the position of the interface, and a1, b1, a2, and b2 are all coefficients. Step 9: Take the absolute value of the coefficient a1 in the first segment of the linear equation curve to represent the velocity index value of the interface descent. The velocity index value is used to compare the results of the settlement detection at different times.
[0020] In step 7 above, during a single observation, the size of N is determined based on the statistical results of the interface height values. If the change in the height values of adjacent interfaces is less than the threshold, the image acquisition for this observation is stopped.
[0021] Figure 3 This represents the image processing result of one of the N images in step 7 during a single observation, which marks the bounding box of the region of interest, as well as the height marker line and height value of the interface.
[0022] Figure 4 To fit the line graph, images were acquired at a fixed time interval of once per second to ensure consistency in the trajectory curve fitting. In the fitted line graph, the x-axis represents the time interval, and the y-axis represents the position of the interface. The absolute value of the coefficient a1 of the first straight line in the fitted two linear equation curves (y = a1*x+b1, y = a2*x+b2) represents the rate of descent of the interface. The coefficient a1 reflects the rate of change of the y-value during descent; a larger absolute value of a1 indicates better settlement effect, and vice versa. No standard value is set for the quality of the observed settlement phenomenon; it is only used for comparison between adjacent observations.
[0023] Figure 5 The fitting process is shown in the figure. Figure 5 In this process, images are acquired at a fixed time interval of once per second to ensure the consistency of the trajectory curve fitting. The descent of the settlement interface is summarized into two stages: the descent stage and the completion stage. A polygonal line composed of two straight lines is fitted, where the slope a1 of the first straight line represents the descent velocity, and the second straight line a2 represents the state of completion of settlement. In the fitted polygonal line graph, the horizontal direction represents time, and the vertical direction represents the position of the interface. Figure 5 Medium-dark green (e.g.) Figure 5 The dark green dot in the upper left corner represents the height of the interface at each moment. After fitting a polygonal line to the dark green dot, the fitted straight line is represented by a light green dot. The absolute value of the coefficient a1 in the fitted polygonal line equation (the slope of the equation) represents the rate of interface descent. The coefficient a1 reflects the rate of change of the y-value in the straight line. The larger the absolute value of a1, the better the settling effect (faster settling speed) in this observation, and vice versa. The observed settling speed can directly reflect the effect of the upstream reagent addition to the slurry.
[0024] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A machine vision-based flocculation and sedimentation observation system, characterized in that: It includes an industrial camera (1), a transparent observation chamber (2) corresponding to the industrial camera (1), and a strip light source (3) set next to the transparent observation chamber (2). The industrial camera (1) is set vertically facing the front of the transparent observation chamber (2), and there is a flocculation solution in the transparent observation chamber (2).
2. The flocculation and sedimentation observation system based on machine vision according to claim 1, characterized in that: The industrial camera (1) is a webcam, and the strip light source (3) is a planar light source.
3. The flocculation and sedimentation observation system based on machine vision according to claim 2, characterized in that: The transparent observation room (2) is a cubic box made of acrylic material.
4. A machine vision-based method for observing flocculation and sedimentation, utilizing the observation system of any one of claims 1 to 3, characterized in that... Includes the following steps: Step 1: The industrial camera (1) is positioned vertically to the side of the transparent observation room (2) to capture an image of the transparent observation room (2); Step 2: Pre-define the region of interest in the image for analysis and calculation; Step 3: Perform grayscale processing on the region of interest in the image; Step 4: Within the region of interest of the image, calculate the average grayscale value of each row of pixels along the column direction and place them in a list; Step 5: Calculate the gradient value of the average grayscale value of all items in the list from Step 4 along the column direction; Step 6: Determine the gradient value of the interface: Take the row with the smallest gradient value to represent the interface of settlement, and record the height mark line and Y coordinate value of the interface. The Y coordinate value represents the height value of the interface. Step 7: Repeat steps 1 to 6 to continuously capture N images of the transparent observation chamber, record the height values of N interfaces, and then proceed to step 8; Step 8: Fit two linear equation curves (y = a1*x+b1, y = a2*x+b2) using N height values to represent the process of the interface descending, where the x-axis is the time interval, the y-axis is the position of the interface, and a1, b1, a2, and b2 are all coefficients. Step 9: Take the absolute value of the coefficient a1 in the first segment of the linear equation curve to represent the velocity index value of the interface descent. The velocity index value is used to compare the results of the settlement detection at different times.
5. The flocculation and sedimentation observation method based on machine vision according to claim 4, characterized in that: In step 2, the region of interest in the image is a rectangular area near the edge of the supplementary lighting side of the strip light source (3) to improve the calculation accuracy.
6. The flocculation and sedimentation observation method based on machine vision according to claim 4, characterized in that: In step 3, grayscale processing is performed using the average value, weighted average method, or maximum value method.
7. The flocculation and sedimentation observation method based on machine vision according to claim 4, characterized in that: In step 7, the time interval for acquiring images of the transparent observation chamber is fixed, and the acquisition is performed at a frequency of once per second.