A dust concentration monitoring data source switching method
Patent Information
- Application Number
- CN202611283333.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-24
- Publication Date
- 2026-10-13
AI Technical Summary
其次，在环境浓度剧烈波动的复杂工业现场，单一浓度仪的读数容易出现饱和或数据失真，缺乏多源数据冗余校验机制，难以保证监测结果的持续可靠性
本申请提供了一种粉尘浓度监测数据源切换方法，实时确定浓度仪的当前读数与当前帧图像的角点数量，通过引入角点数量与灰度值对监测数据源进行联合判定与切换，解决了现有粉尘浓度监测设备在极端工况下易发生量程饱和以及误差过大的技术痛点；在工艺生产时，实时监测浓度仪读数，当浓度仪读数大于预设浓度阈值时，对第一切换条件进行判定，即当角点数量较多且灰度值降低时，判定环境中的粉尘浓度发生实质跃升，基于视觉双特征判定条件，将监测数据源切换至摄像机；当角点数量下降且灰度值到达适当范围时，监测数据源重新切换至浓度仪。本申请有效消除了单一设备高浓度饱和带来的测量盲区，保证了粉尘浓度的测量连续性，提高了抗干扰能力与全局数据精度。
Smart Images
Figure CN122901035A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of dust concentration monitoring, and in particular to a method for switching data sources for dust concentration monitoring. Background Technology
[0002] In industrial production and dust monitoring, accurate monitoring of high-concentration dust is crucial for environmental safety and process control. However, current monitoring solutions face numerous challenges in practical applications. First, in dust monitoring practice, conventional physical concentration meters offer extremely high detection accuracy and linearity in low-concentration environments, while camera-based image prediction methods are susceptible to background noise interference in low-concentration ranges, resulting in relatively limited accuracy. Therefore, accurate measurement at low concentrations relies on concentration meters. However, the physical measurement range of mainstream dust monitoring equipment on the market is typically limited to 1 mg / L. When the ambient concentration exceeds this range and enters the high-concentration range, the concentration meter may fail to provide accurate values due to signal saturation. In this case, image feature analysis using cameras becomes an effective supplementary means of predicting high-concentration values. Therefore, establishing a scientific switching strategy to achieve rapid and accurate switching between conventional concentration meters (accurate low-concentration monitoring) and camera prediction results (high-concentration range compensation) is a pressing technical challenge. Second, in complex industrial environments with drastic fluctuations in ambient concentration, readings from a single concentration meter are prone to saturation or data distortion. The lack of multi-source data redundancy verification mechanisms makes it difficult to guarantee the continuous reliability of monitoring results.
[0003] Therefore, based on the above problems, there is an urgent need to provide a method for switching data sources for dust concentration monitoring. This method can achieve automatic switching of output sources through rigorous logical judgment, ensuring monitoring accuracy and response speed in high-concentration environments, eliminating measurement blind spots caused by concentration saturation of a single device, and ensuring the continuity of dust concentration measurement. Summary of the Invention
[0004] The purpose of this application is to provide a method for switching data sources for dust concentration monitoring, which can eliminate the measurement blind zone caused by the concentration saturation of a single device, ensure the continuity of dust concentration measurement, and improve measurement accuracy.
[0005] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a method for switching dust concentration monitoring data sources, including: Acquire the baseline grayscale value of the target area image; and acquire the current reading of the concentration meter and the current image frame captured by the camera in real time during the production process; The number of corner points in the current image frame is determined by using an accelerated segmented test feature algorithm. When the current reading of the concentration meter is less than or equal to the preset concentration threshold, the monitoring data source is the concentration meter; When the current reading of the concentration meter is greater than the preset concentration threshold, determine the grayscale value of the current image frame; Determine whether the number of corner points and the grayscale value meet the first switching condition; when the number of corner points and the grayscale value meet the first switching condition, switch the monitoring data source from the concentration meter to the camera; and use the grayscale value of the current image frame as the reference grayscale value; the first switching condition is that the number of corner points is greater than the reference number of corner points and the grayscale value is less than the baseline grayscale value; the reference number of corner points is determined according to a preset concentration threshold; The grayscale value and the number of corner points of the current image frame are continuously acquired. When the number of corner points and the grayscale value meet the second switching condition, the monitoring data source is switched from the camera to the density meter. The second switching condition is that the number of corner points is less than the maximum number of corner points and the grayscale value is greater than the reference grayscale value and the magnification grayscale value. The magnification grayscale value is less than the reference grayscale value and the magnification grayscale value is determined according to the reference grayscale value.
[0006] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a method for switching data sources for dust concentration monitoring. It determines the current reading of the concentration meter and the number of corner points in the current frame image in real time. By introducing the number of corner points and grayscale values to jointly determine and switch the monitoring data source, it solves the technical pain points of existing dust concentration monitoring equipment, which is prone to range saturation and excessive errors under extreme operating conditions. During production, the concentration meter reading is monitored in real time. When the concentration meter reading exceeds a preset concentration threshold, a first switching condition is determined: when the number of corner points is large and the grayscale value decreases, it is determined that the dust concentration in the environment has substantially increased. Based on the visual dual-feature determination condition, the monitoring data source is switched to the camera. When the number of corner points decreases and the grayscale value reaches an appropriate range, the monitoring data source is switched back to the concentration meter. This application effectively eliminates the measurement blind zone caused by high concentration saturation of a single device, ensures the continuity of dust concentration measurement, and improves anti-interference capability and overall data accuracy. Attached Figure Description
[0007] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0008] Figure 1 This is a flowchart illustrating a dust concentration monitoring data source switching method in one embodiment of this application; Figure 2 This is a flowchart illustrating a dust concentration monitoring data source switching method in one embodiment of this application. Detailed Implementation
[0009] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0010] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0011] In one exemplary embodiment, such as Figure 1 and Figure 2 As shown, a method for switching data sources for dust concentration monitoring is provided, including the following steps S1 to S6. Wherein: S1: Obtain the baseline grayscale value of the target area image; and obtain the current reading of the concentration meter and the current image frame captured by the camera during the production process in real time.
[0012] S1 specifically includes: S11: Before the production process, use a camera to collect image sequences of the target area; and determine the grayscale fluctuation range of the image sequence within a continuous time sliding window.
[0013] Specifically, after the equipment powers on and completes its self-test, before the formal start of the production process, the front-end camera starts continuously acquiring video streams (image sequences). Within a continuous 0.5s time window, the grayscale fluctuation of the global image or the region of interest (ROI) of the image target is extremely poor.
[0014] S12: When the grayscale fluctuation range is less than or equal to the preset range threshold, the average grayscale value of the image within the continuous time sliding window is used as the reference grayscale value.
[0015] Specifically, if the grayscale fluctuation range within the 0.5s time window is determined to be ≤5 (preset range threshold), then the current environment is determined to be in a steady state, and the average grayscale value of the entire image at this time is recorded as the baseline grayscale value.
[0016] S13: When the grayscale fluctuation range is greater than the preset range threshold within a preset time, the grayscale value of the last frame image within the preset time is used as the reference grayscale value.
[0017] Specifically, if the image capture fails to meet the steady-state fluctuation condition due to ambient light disturbances, forced convergence will be triggered. The grayscale value of the last valid image in the 3rd second (preset time) will be directly extracted as the reference grayscale value to enter the standby state on time.
[0018] S14: After the process begins, the current reading of the concentration meter and the current image frame captured by the camera are acquired in real time.
[0019] S2: Use the Accelerated Segment Test (FAST) feature algorithm to determine the number of corner points in the current image frame.
[0020] S2 specifically includes: S21: Select the target pixel in the current image frame.
[0021] Specifically, the target pixel is selected from the current image frame acquired in real time.
[0022] S22: Determine multiple adjacent pixels with the target pixel as the center and a preset pixel range as the radius.
[0023] In one exemplary embodiment, 16 adjacent pixels are extracted from a discrete circle (Bresenham circle) centered on the target pixel and with a radius of a preset pixel range (e.g., 3).
[0024] S23: Based on the gray values of adjacent pixels, classify adjacent pixels according to the state classification function to obtain the classification result.
[0025] Set a grayscale tolerance threshold and define the first grayscale value on the discrete circle. FAST defines a state classification function for each pixel relative to the target pixel, and classifies neighboring pixels based on this function to obtain classification results. These results include dark feature points, bright feature points, and similar feature points. The state classification function for neighboring pixels is defined as follows: ; Among them, among them, For the first A classification function of the state of each neighboring pixel relative to the target pixel. A dark feature point indicates that the adjacent pixel is significantly darker than the target pixel. For the first The grayscale values of adjacent pixels The grayscale value of the target pixel. This refers to the gray-scale difference threshold for FAST corner detection, or simply the gray-scale difference threshold. For similar feature points, A bright feature point indicates that the adjacent pixel is significantly brighter than the target pixel. For the first The nth adjacent pixel, that is, the nth pixel on the Bresenham circle with radius 3 centered at the target pixel. Each pixel.
[0026] S24: Based on the classification results, the target pixels are filtered using the accelerated detection function to obtain candidate corner points.
[0027] To reduce computational load, an accelerated detection function is defined before determining whether a target pixel is an initial corner point. This function pre-tests four adjacent pixels along the cross direction of the discrete circumference to filter target pixels. The accelerated detection function is calculated as follows: ; in, To accelerate the detection function, For indicator functions, For the first A classification function of the state of each neighboring pixel relative to the target pixel. For dark feature points, These are bright feature points.
[0028] S25: Based on the candidate corner points, determine the initial corner points according to the corner point determination equation.
[0029] The necessary and sufficient condition for determining a target pixel as a feature corner point is: there exists a continuous feature corner point on the discrete circumference. There are 100 pixels, and their classification results belong to 1000 pixels. (Dark feature points), or simultaneously belong to (Bright feature points). That is, if and only if there are continuous features on the discrete circle. (preferred) The gray values of adjacent pixels are strictly greater than 1. or strictly less than At that time, the judgment point Let be the initial corner point. The equation for determining the initial corner point is as follows: ; in, The initial corner point determination result; if the condition is met, then the target pixel... As the initial corner point, Numbering the starting point of a continuous pixel segment, indicating the starting point from the [number of pixels]. Start by judging from the points on the circumference. To start from the starting point The offset number of the starting adjacent pixel, This refers to the number of consecutive adjacent pixels that meet the condition, such as 9 consecutive or 12 consecutive adjacent pixels being both bright and dark feature points. For the first A classification function of the state of each neighboring pixel relative to the target pixel. For dark feature points, For bright feature points, For the first Adjacent pixels, , The remainder is used to implement the cyclic numbering of 16 adjacent pixels.
[0030] This application avoids complex image gradient differentiation operations and achieves microsecond-level feature point response, providing high real-time data support for subsequent extraction of the total number of corner points reflecting dust concentration and environmental dirtiness.
[0031] S26: Based on the Non-Maximum Suppression (NMS) method, determine the response score corresponding to the initial corner point; and determine the corner point based on the response score.
[0032] To address the corner clustering phenomenon that may occur due to dense noise in high-concentration dust detection, non-maximum suppression needs to be introduced. The response score of the initially identified corner points is then calculated to further confirm their location. The formula for calculating the response score is as follows: ; in, Initial corner point The response score, For the first The grayscale values of adjacent pixels The grayscale value of the target pixel. The grayscale difference threshold. The set of neighboring pixels that are classified as bright feature points. It is the set of neighboring pixels that are classified as dark feature points.
[0033] The sensitivity threshold for corner detection is finely adjusted downwards. When the threshold is lowered to <3, the total number of corners (effective corners) in the current image frame is counted, and it is determined whether the number of corners is greater than a preset corner number threshold (e.g., 10). If the number of corners is greater than the preset threshold (>10), the current image frame is denoised using a Gaussian smoothing filter to remove high-frequency pseudo-random noise. The current image frame is then re-acquired, and the number of corners is determined again. If the number of corners determined again is greater than the preset threshold (>10), an abnormal state of physical contamination of the lens or background is triggered. Data output is immediately intercepted, and a hardware maintenance warning is generated, prompting the cleaning of the lens or replacement of the background. After the physical contamination is removed, the number of corners is determined again using an accelerated segmented test feature algorithm. If the number of corners is less than or equal to the preset threshold (≤10), it is determined whether the concentration meter reading is greater than a preset concentration threshold.
[0034] S3: When the current reading of the concentration meter is less than or equal to the preset concentration threshold, the monitoring data source is the concentration meter.
[0035] In an exemplary embodiment, when the current reading of the concentration meter is less than or equal to a preset concentration threshold (≤0.9), the monitoring data source is the concentration meter, the concentration meter reading A is directly output, and A is used as the current dust concentration.
[0036] S4: When the current reading of the concentration meter is greater than the preset concentration threshold, determine the grayscale value of the current image frame.
[0037] In one exemplary embodiment, when the current reading of the concentration meter is greater than a preset concentration threshold (>0.9), the grayscale value of the current image frame is calculated.
[0038] S5: Determine whether the number of corner points and grayscale value meet the first switching condition; when the number of corner points and grayscale value meet the first switching condition, switch the monitoring data source from the concentration meter to the camera.
[0039] The first switching condition is that the number of corner points is greater than the number of reference corner points and the grayscale value is less than the baseline grayscale value; the number of reference corner points is determined based on a preset concentration threshold. The formula for the first switching condition is as follows: ; ; in, This represents the number of corner points in the current image frame. For reference, the number of corner points , The baseline corner point count is the number of corner points corresponding to the current reading of the concentration meter equaling the preset concentration threshold. To minimize the number of corner points, The grayscale value of the current image frame. This is the baseline grayscale value.
[0040] When the number of corner points and the grayscale value meet the first switching condition, the monitoring data source is switched from the concentration meter to the camera, and the camera reading B is used as the current dust concentration, while the grayscale value of the current image frame is used as the reference grayscale value. The number of corner points in the current image frame is recorded as... .
[0041] S6: Continuously acquire the grayscale value and corner count of the current image frame. When the corner count and grayscale value meet the second switching condition, switch the monitoring data source from the camera to the concentration meter.
[0042] The second switching condition is that the number of corner points is less than the maximum number of corner points and the grayscale value is greater than the reference grayscale value and the magnification grayscale value. The magnification grayscale value is less than the reference grayscale value and the magnification grayscale value is determined based on the baseline grayscale value.
[0043] Specifically, while outputting the camera reading B, the grayscale value of the current image frame continues to be calculated in real time. With the number of corner points It then determines whether the number of corner points and the grayscale value meet the second switching condition. The formula for calculating the second switching condition is as follows: ; ; in, The maximum number of corner points. , For reference grayscale values, This is the grayscale value at magnification. , As the baseline grayscale value, The number of corner points required to satisfy the first switching condition.
[0044] In an exemplary embodiment, when determining whether the number of corner points and the grayscale value meet the second switching condition, the first step is to determine... Whether it is true; if If this is not established, then the monitoring data source will remain the camera; if If true, then further judgment is needed. and Whether it is true; if and If this is not established, then the monitoring data source will remain the camera; if and If both conditions are met, the monitoring data source will be switched from the camera to the concentration meter.
[0045] When the grayscale value of the current image frame With the number of corner points If the second switching condition is not met, the monitoring data source is the camera, and the camera reading B is continuously output. When the grayscale value of the current image frame... With the number of corner points When the second switching condition is met, the monitoring data source is switched from the camera to the concentration meter, the concentration meter reading A is output, and the previously determined number of reference corner points and the number of maximum corner points are cleared.
[0046] This application dynamically captures images and calculates grayscale fluctuations within 0.5 seconds before production begins. This mechanism allows the application to extract baseline grayscale values based on the actual lighting and environment each time it starts, rather than relying on rigid factory-preset parameters, thus improving its scene adaptability. Secondly, this application uses the FAST corner detection algorithm, which is known for its extremely fast feature extraction speed and has lower computational complexity compared to conventional feature point algorithms. Applying it to real-time frame-by-frame detection greatly reduces the computational load, ensuring millisecond-level instantaneous responses to environmental changes and pollution conditions during continuous production, meeting the stringent real-time monitoring requirements of industrial sites. Thirdly, this application does not rigidly rely on a single sensor, but combines the concentration meter reading A with image visual features (number of corner points and image grayscale). When the concentration meter reading is abnormal or reaches a specific threshold (A>0.9), image features (a sharp increase in corner points and a decrease in grayscale) are automatically introduced to determine whether to switch the output source, ensuring the most reliable readings are still output even in complex production environments.
[0047] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0048] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.
Claims
1. A method for switching data sources for dust concentration monitoring, characterized in that, The method includes: Acquire the baseline grayscale value of the target area image; and acquire the current reading of the concentration meter and the current image frame captured by the camera in real time during the production process; The number of corner points in the current image frame is determined by using an accelerated segmented test feature algorithm. When the current reading of the concentration meter is less than or equal to the preset concentration threshold, the monitoring data source is the concentration meter; When the current reading of the concentration meter is greater than the preset concentration threshold, determine the grayscale value of the current image frame; Determine whether the number of corner points and the grayscale value meet the first switching condition; when the number of corner points and the grayscale value meet the first switching condition, switch the monitoring data source from the concentration meter to the camera; and use the grayscale value of the current image frame as the reference grayscale value; the first switching condition is that the number of corner points is greater than the reference number of corner points and the grayscale value is less than the baseline grayscale value; the reference number of corner points is determined according to a preset concentration threshold; The grayscale value and the number of corner points of the current image frame are continuously acquired. When the number of corner points and the grayscale value meet the second switching condition, the monitoring data source is switched from the camera to the density meter. The second switching condition is that the number of corner points is less than the maximum number of corner points and the grayscale value is greater than the reference grayscale value and the magnification grayscale value. The magnification grayscale value is less than the reference grayscale value and the magnification grayscale value is determined according to the reference grayscale value.
2. The dust concentration monitoring data source switching method according to claim 1, characterized in that, The acquisition of the reference grayscale value of the target region image specifically includes: Before the production process, an image sequence of the target area is acquired using a camera; and the grayscale fluctuation range of the images within a continuous time window is determined. When the grayscale fluctuation range is less than or equal to the preset range threshold, the average grayscale value of the image sequence within the continuous time sliding window is used as the reference grayscale value. If the grayscale fluctuation range is greater than the preset range threshold within a preset time, the grayscale value of the last frame image within the preset time is used as the reference grayscale value.
3. The dust concentration monitoring data source switching method according to claim 1, characterized in that, The method of using an accelerated segmented test feature algorithm to determine the number of corner points in the current image frame specifically includes: Select the target pixel in the current image frame; With the target pixel as the center and a preset pixel range as the radius, determine multiple adjacent pixels; Based on the gray values of adjacent pixels, adjacent pixels are classified according to a state classification function to obtain classification results; the classification results include: dark feature points, bright feature points, and similar feature points; Based on the classification results, and using an accelerated detection function, the target pixels are filtered to obtain candidate corner points; Based on the candidate corner points, the initial corner points are determined according to the corner point determination equation; Based on the nonmaximum suppression method, the response score corresponding to the initial corner point is determined; and the corner point is determined based on the response score.
4. The dust concentration monitoring data source switching method according to claim 3, characterized in that, The step of classifying adjacent pixels based on their grayscale values using a state classification function to obtain classification results specifically includes: Using formula Classify adjacent pixels; in, For the first A classification function of the state of each neighboring pixel relative to the target pixel. For dark feature points, For the first The grayscale values of adjacent pixels The grayscale value of the target pixel. The grayscale difference threshold. For similar feature points, These are bright feature points.
5. The dust concentration monitoring data source switching method according to claim 3, characterized in that, The step of filtering target pixels based on the classification results and using an accelerated detection function to obtain candidate corner points specifically includes: Using formula Filter the target pixels; in, To accelerate the detection function, For indicator functions, For the first A classification function of the state of each neighboring pixel relative to the target pixel. For dark feature points, These are bright feature points.
6. The dust concentration monitoring data source switching method according to claim 3, characterized in that, The step of determining the initial corner point based on the candidate corner points and the corner point determination equation specifically includes: Using formula Determine the initial corner points; in, This is the initial corner determination result. Numbering the starting point of a continuous pixel segment. To start from the starting point The offset number of the starting adjacent pixel, This refers to the number of consecutive adjacent pixels that satisfy the condition. For the first A classification function of the state of each neighboring pixel relative to the target pixel. For dark feature points, For bright feature points, For the first Adjacent pixels, , To take the remainder when divided by 16.
7. The dust concentration monitoring data source switching method according to claim 3, characterized in that, The non-maximum suppression method for determining the response score corresponding to the initial corner point specifically includes: Using formula Determine the response score corresponding to the initial corner point; in, Initial corner point The response score, For the first The grayscale values of adjacent pixels The grayscale value of the target pixel. The grayscale difference threshold. The set of neighboring pixels that are classified as bright feature points. It is the set of neighboring pixels that are classified as dark feature points.
8. The dust concentration monitoring data source switching method according to claim 1, characterized in that, The method of using an accelerated segmented test feature algorithm to determine the number of corner points in the current image frame also includes: If the number of corner points is greater than a preset corner point number threshold, the current image frame is denoised using a Gaussian smoothing filter, and the number of corner points is determined again. If the number of corner points determined again is greater than the preset corner point number threshold, an early warning is issued. If the number of corner points is less than or equal to the preset corner point number threshold, the concentration meter reading is determined to be greater than a preset concentration threshold.
9. The dust concentration monitoring data source switching method according to claim 1, characterized in that, The process of determining the number of reference corner points includes: Use the number of corner points corresponding to the preset concentration threshold as the baseline number of corner points; Using formula Determine the number of reference corner points; among which, For reference, the number of corner points As a reference number of corner points, This represents the minimum number of corner points.
10. The dust concentration monitoring data source switching method according to claim 1, characterized in that, The process of determining the magnification grayscale value includes: Using formula Determine the grayscale value at the magnification; in, This is the grayscale value at magnification. This is the baseline grayscale value.