Dust image acquisition equipment based on uniformity detection
By designing a dust image acquisition device based on uniformity detection, the problem of inaccurate dust image acquisition and data calibration in the existing technology is solved, achieving high precision and system applicability in dust concentration detection, and providing uniformly distributed training data.
Patent Information
- Application Number
- CN202512024107.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-30
- Publication Date
- 2026-03-03
AI Technical Summary
Existing dust image acquisition and data labeling technologies suffer from spatial mismatch between "points" and "areas," unreliable and inaccurate label data, and large model training errors, resulting in low accuracy of vision-based dust concentration detection.
A dust image acquisition device based on uniformity detection was designed, including a system control module, a dust generation module, an image acquisition module, an air source processing module, and a dust collection module. Through components such as a fluidized bed, a mixing chamber, and an air pump, the device ensures that the dust is uniformly distributed in the observation chamber, and captures images through a dust concentration meter and a camera to provide accurate training data.
It improves the accuracy of dust concentration detection and the applicability of the system in different scenarios, ensures the uniformity of dust distribution, provides high-precision training data, and supports high-precision detection by visual models.
Smart Images
Figure CN121595412A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of dust monitoring, and specifically to a dust image acquisition device based on uniformity detection. Background Technique
[0002] Industrial and mining enterprises (such as coal mines, cement plants, construction sites, etc.) inevitably generate a large amount of dust during the production process. Among them, high-concentration dust can cause explosion accidents under specific conditions, resulting in huge economic losses and casualties. Therefore, real-time and accurate monitoring of dust concentration is an important guarantee for safe production.
[0003] In recent years, with the rapid development of deep learning and machine vision technologies, dust concentration detection methods based on image processing have gradually become a research hotspot. However, constructing a high-precision visual dust detection model highly depends on a large amount of accurately labeled training data (i.e., "image-concentration" pairs).
[0004] In the existing dust image acquisition and data calibration technologies, there are significant defects such as the spatial mismatch problem between "points" and "surfaces", untrue and unreliable label data, and large model training errors. Therefore, constructing an acquisition system that can ensure the uniform distribution of dust in the acquisition space, thereby ensuring the strict correspondence between sensor readings and image features, has become the key technical bottleneck for improving the accuracy of vision-based dust concentration detection. Summary of the Invention
[0005] The purpose of the present invention is to provide a dust image acquisition device based on uniformity detection, which can generate uniformly distributed dust in the mixing chamber, evaluate the dispersion degree of concentration data, judge the current dust distribution state, effectively ensure that the uniformity of dust concentration distribution is within the required range, and improve the accuracy of the measured concentration data, thus solving the problems raised in the above background technique.
[0006] To achieve the above purpose, the present invention provides the following technical solutions:
[0007] A dust image acquisition device based on uniformity detection includes a system control module, a dust generation module, an image acquisition module, an air source processing module, and a dust collection module arranged on a frame. The dust generation module is used to load powder and start when the dust concentration collected by the image acquisition module is 0, and upload the concentration data to the system control module.
[0008] The system control module is used to set the dust concentration C0 and the dispersion coefficient CV0, calculate the dispersion degree of the dust, and when the dispersion coefficient CV < CV0 and the dust concentration C = C0, control the image acquisition module to start collecting images.
[0009] Preferably, the system control module includes a host computer and a display, which are located on the side end face of the rack, with the display located above the host computer.
[0010] Preferably, the dust generating module includes a fluidized bed, a concentration meter, and a mixing chamber. The mixing chamber is located at the upper end of the frame, and the fluidized bed is installed on the mixing chamber and connected to the mixing chamber via an elbow.
[0011] Preferably, the image acquisition module includes a camera, a background panel, an observation chamber, and a ring light source. The background panel is installed inside the observation chamber, the camera is installed on the observation chamber, and the ring light source is installed on the top of the observation chamber.
[0012] Preferably, the air source processing module includes an air pump and a dryer, which are located on the side end face of the frame, with the dryer located above the air pump.
[0013] Preferably, the dust collection module includes a cone collector and a dust filter, the dust filter being located at the lower end of the frame, and the front end of the dust filter being connected to the cone collector.
[0014] Preferably, the observation chamber is equipped with a dust concentration sampling point, which is connected to a dilution chamber to dilute the dust concentration by a specific factor, so that the dust concentration in the dilution chamber is less than 1000 mg / m³. Then, a dust concentration meter is connected to the dilution chamber to accurately measure the high dust concentration.
[0015] Preferably, the host computer is set with dust concentration C0 and dispersion coefficient CV0;
[0016] Start the dust collection device and dust concentration meter. When the average dust concentration C in the observation chamber is greater than 0, increase the dust filter until the average dust concentration in the observation chamber is 0.
[0017] The fluidized bed is loaded with powder, the ring light source is turned on, the air pump is started, the fluidized bed begins to generate powder, and the uniformity of the dust is adjusted by the turbulence in the mixing chamber so that the uniformity of the dust reaches the observation chamber to meet the requirements.
[0018] The dust uniformity detection is determined by the dispersion of concentration data uploaded by dust concentration meters at the four dust concentration sampling points. The dispersion calculation method is as follows:
[0019]
[0020] in, The population standard deviation is 1. The population mean and The mathematical expression is as follows:
[0021]
[0022] in, The number of dust concentration data points. This represents the i-th dust concentration data value;
[0023] When the dust concentration dispersion CV calculated by the host computer is greater than CV0, it indicates that the dust concentration uniformity has not met the required standards. Adjust the turbulence state and intensity in the mixing chamber until the dust concentration dispersion CV is reduced. <CV0;
[0024] The host computer calculates the average concentration C of the four dust concentration meters. If C > C0, the air pump flow rate is reduced to decrease the amount of dust generated in the fluidized bed; if C < C0, the air pump flow rate is increased to increase the amount of dust generated in the fluidized bed.
[0025] Each time the fluidized bed flow rate is adjusted, the dispersion coefficient needs to be re-detected. When C=C0, the camera starts capturing dust images.
[0026] Compared with the prior art, the beneficial effects of the present invention are:
[0027] This invention constructs a dust image acquisition device based on uniformity detection. It can generate uniformly distributed dust in the mixing chamber, providing accurate training data for the dust detection model. It evaluates the dispersion of concentration data in the observation chamber to determine the current dust distribution status, effectively ensuring that the uniformity of dust concentration distribution is within the required range and guaranteeing the applicability of the system in different scenarios. A quick-release structure for the camera and background is designed, allowing users to easily replace the camera and background. A dilution chamber is connected outside the dust concentration sampling point to dilute the dust concentration by a certain factor, so that the dust concentration in the dilution chamber is between 0-1000 mg / m³, improving the accuracy of the measured concentration data. Attached Figure Description
[0028] Figure 1 This is a schematic diagram of the dust image acquisition device based on uniformity detection according to the present invention;
[0029] Figure 2 This is a side view of the dust image acquisition device based on uniformity detection according to the present invention.
[0030] Figure 3 This is a block diagram of the dust image acquisition device based on uniformity detection according to the present invention.
[0031] Figure 4 This is a flowchart of the dust image acquisition device based on uniformity detection according to the present invention.
[0032] In the figure: 11, host computer; 12, display; 21, fluidized bed; 23, mixing chamber; 31, camera; 32, background board; 33, observation chamber; 331, dust concentration sampling point; 34, annular light source; 41, air pump; 42, dryer; 51, conical collector; 52, dust filter. Specific implementation mode
[0033] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.
[0034] In order to solve the significant defects such as the spatial mismatch problem between "points" and "planes", untrue and unreliable label data, and large model training errors in the existing dust image acquisition and data calibration technologies, resulting in low detection accuracy of dust concentration based on vision, please refer to Figures 1-2 , the present embodiment provides the following technical solutions:
[0035] A dust image acquisition device based on uniformity detection, including a system control module, a dust generation module, an image acquisition module, a gas source processing module and a dust collection module arranged on a frame;
[0036] In this embodiment, the system control module includes a host computer 11 and a display 12. The host computer 11 and the display 12 are located on the side end face of the frame, and the display 12 is located above the host computer 11.
[0037] It should be noted that the dust concentration C0 and the dispersion coefficient CV0 are set through the host computer 11. When the dust concentration dispersion degree CV calculated by the host computer 11 is greater than CV0, it means that the dust concentration uniformity does not meet the required standard. Adjust the turbulence state and intensity of the mixing chamber 23 until the dust concentration dispersion degree CV is less than CV0; at the same time, the host computer 11 calculates the average concentration C of the four dust concentration meters. If C > C0, reduce the flow rate of the air pump 41 to reduce the powder generation amount of the fluidized bed 21; if C < C0, increase the flow rate of the air pump 41 to increase the powder generation amount of the fluidized bed 21; after each adjustment of the flow rate of the fluidized bed 21, it is necessary to re-detect the dispersion coefficient. When C = C0, the camera 31 starts to capture dust images.
[0038] It should be noted that it is displayed in real time through the display 12 for the convenience of the operator to view.
[0039] In this embodiment, the dust generation module includes a fluidized bed 21, a concentration meter, and a mixing chamber 23. The mixing chamber 23 is located at the upper end of the frame, and the fluidized bed 21 is disposed on the mixing chamber 23. The fluidized bed 21 is connected to the mixing chamber 23 through an elbow.
[0040] It should be noted that a fluidized bed is a reactor that uses gas or liquid to suspend solid particles to achieve gas-solid or liquid-solid reactions. It includes types such as dispersed fluidized beds and cohesive fluidized beds. Its critical fluidization velocity is the minimum linear velocity required for the fluid to carry the particles.
[0041] In this embodiment, the fluidized bed 21 is loaded with powder, the ring light source 34 is turned on, and the air pump 41 is started. The fluidized bed 21 begins to generate powder, and the uniformity of the dust is adjusted by the turbulence of the mixing chamber 23 so that the uniformity of the dust reaches the observation chamber 33 and meets the requirements.
[0042] It should be noted that dust concentration is monitored using a concentration meter, and the uniformity of dust is detected based on the monitored dust concentration data, which in turn determines the degree of dispersion. The method for calculating the dispersion is as follows:
[0043]
[0044] in, The population standard deviation is 1. The population mean and The mathematical expression is as follows:
[0045]
[0046] in, The number of dust concentration data points. Let be the i-th dust concentration data value.
[0047] It should be noted that the uniformity of the dust is adjusted by turbulence in the mixing chamber 23 so that the uniformity of the dust reaches the observation chamber 33 and meets the requirements.
[0048] Specifically, it can generate uniformly distributed dust in the mixing chamber 23, providing accurate training data, i.e., "image-concentration" pairs, for the dust detection model; it can also evaluate the dispersion of the concentration data in the mixing chamber 23 to determine the current dust distribution status, effectively ensuring that the uniformity of the dust concentration distribution is within the required range, and ensuring the applicability of the system in different scenarios.
[0049] In this embodiment, the image acquisition module includes a camera 31, a background plate 32, an observation chamber 33, and a ring light source 34. The background plate 32 is arranged inside the observation chamber 33, the camera 31 is arranged on the observation chamber 33, and the ring light source 34 is arranged on the top of the observation chamber 33.
[0050] In this embodiment, dust images are captured by camera 31.
[0051] It should be noted that the camera 31 and the backdrop 32 adopt a quick-release structure, which allows users to easily change the model of the camera 31 and the backdrop 32, making it convenient to use.
[0052] In this embodiment, the air source processing module includes an air pump 41 and a dryer 42, which are located on the side end face of the frame, with the dryer 42 located above the air pump 41.
[0053] In this embodiment, the dust collection module includes a conical collector 51 and a dust filter 52. The dust filter 52 is located at the lower end of the frame, and the front end of the dust filter 52 is connected to the conical collector 51.
[0054] In this embodiment, since the dust concentration instrument has high accuracy and fast response time in the measurement range of 0-1000mg / m³, a dust concentration sampling point 331 is set on the observation chamber 33, and the dust concentration sampling point 331 is connected to a dilution chamber to dilute the dust concentration by a specific factor, so that the dust concentration in the dilution chamber is lower than 1000mg / m³. Then, the dust concentration meter is connected to the dilution chamber to ensure that the dust concentration meter can accurately measure high dust concentrations.
[0055] Specifically, a dilution chamber is connected to the dust concentration sampling point 331 to dilute the dust concentration by a certain factor, so that the dust concentration in the dilution chamber is between 0-1000 mg / m³, thereby improving the accuracy of the measured concentration data.
[0056] It should be noted that, in order to better demonstrate the image acquisition process for uniformly distributed dust, this embodiment now provides a method for acquiring images of uniformly distributed dust, including:
[0057] Set the dust concentration C0 and the coefficient of variation CV0 on the host computer 11;
[0058] Start the dust collection device and dust concentration meter. When the average dust concentration C in the observation chamber 33 is greater than 0, increase the dust filter 52 until the average dust concentration in the observation chamber 33 is 0.
[0059] Powder is loaded into the fluidized bed 21, the ring light source 34 is turned on, and the air pump 41 is started. The fluidized bed 21 begins to generate powder, and the uniformity of the dust is adjusted by the turbulence of the mixing chamber 23 so that the uniformity of the dust reaches the observation chamber 33 and meets the requirements.
[0060] The dust uniformity detection is determined by the dispersion of concentration data uploaded from dust concentration meters at four dust concentration sampling points 331. The dispersion calculation method is as follows:
[0061]
[0062] in, The population standard deviation is 1. The population mean and The mathematical expression is as follows:
[0063]
[0064] in, The number of dust concentration data points. This represents the i-th dust concentration data value;
[0065] When the dust concentration dispersion CV calculated by the host computer 11 is greater than CV0, it indicates that the dust concentration uniformity has not met the required standard. The turbulence state and intensity of the mixing chamber 23 are adjusted until the dust concentration dispersion CV... <CV0;
[0066] The host computer 11 calculates the average concentration C of the four dust concentration meters. If C > C0, the flow rate of the air pump 41 is reduced, so that the amount of dust generated by the fluidized bed 21 is reduced; if C < C0, the flow rate of the air pump 41 is increased, so that the amount of dust generated by the fluidized bed 21 is increased.
[0067] Each time the flow rate of fluidized bed 21 is adjusted, the dispersion coefficient needs to be re-detected. When C=C0, camera 31 starts to capture dust images.
[0068] In summary, a dust image acquisition device based on uniformity detection was constructed, which can generate uniformly distributed dust in the mixing chamber 23, providing accurate training data, i.e., "image-concentration" pairs, for the dust detection model. The dispersion of the concentration data in the mixing chamber 23 is evaluated to determine the current dust distribution state, effectively ensuring that the uniformity of the dust concentration distribution is within the required range, thus guaranteeing the applicability of the system in different scenarios. A quick-release structure for the camera 31 and the background plate 32 is designed, allowing users to easily replace the camera 31 and the background plate 32. A dilution chamber is connected to the dust concentration sampling point 331 to dilute the dust concentration by a certain factor, ensuring that the dust concentration in the dilution chamber is between 0-1000 mg / m³, thereby improving the accuracy of the measured concentration data.
[0069] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0070] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A dust image acquisition device based on uniformity detection, comprising a system control module, a dust generation module, an image acquisition module, an air source processing module, and a dust collection module mounted on a frame, characterized in that, The dust generation module is used to load powder and start when the dust concentration collected by the image acquisition module is 0, and upload the concentration data to the system control module; The system control module is used to set the dust concentration C0 and the dispersion coefficient CV0, calculate the dispersion degree of the dust, and when the dispersion coefficient CV < CV0 and the dust concentration C = C0, control the image acquisition module to start collecting images.
2. The dust image acquisition device based on uniformity detection according to claim 1, characterized in that, The system control module includes a host computer (11) and a display (12). The host computer (11) and the display (12) are located on the side end face of the rack, and the display (12) is located above the host computer (11).
3. The dust image acquisition device based on uniformity detection according to claim 2, characterized in that, The dust generation module includes a fluidized bed (21), a concentration meter and a mixing chamber (23). The mixing chamber (23) is located at the upper end of the rack. The fluidized bed (21) is arranged on the mixing chamber (23), and the fluidized bed (21) is connected to the mixing chamber (23) through an elbow.
4. The dust image acquisition device based on uniformity detection according to claim 3, characterized in that, The image acquisition module includes a camera (31), a background board (32), an observation chamber (33) and an annular light source (34). The background board (32) is arranged inside the observation chamber (33). The camera (31) is arranged on the observation chamber (33), and the annular light source (34) is arranged on the top of the observation chamber (33).
5. The dust image acquisition device based on uniformity detection according to claim 4, characterized in that, The gas source treatment module includes an air pump (41) and a dryer (42). The air pump (41) and the dryer (42) are located on the side end face of the rack, and the dryer (42) is located above the air pump (41).
6. The dust image acquisition device based on uniformity detection according to claim 5, characterized in that, The dust collection module includes a conical collector (51) and a dust filter (52). The dust filter (52) is located at the lower end of the rack, and the front end of the dust filter (52) is connected to the conical collector (51).
7. The dust image acquisition device based on uniformity detection according to claim 6, characterized in that, A dust concentration sampling point (331) is arranged on the observation chamber (33). The dust concentration sampling point (331) is externally connected to a dilution chamber, which is used to dilute the dust concentration by a specific multiple so that the concentration of the dust in the dilution chamber is lower than 1000 mg / m³. Then, a dust concentration meter is connected into the dilution chamber to accurately measure the high dust concentration based on the dust concentration meter.
8. The dust image acquisition device based on uniformity detection according to claim 7, characterized in that, The host computer (11) sets the dust concentration C0 and the dispersion coefficient CV0; Start the dust collection device and the dust concentration meter. When the average dust concentration C in the observation chamber (33) > 0, adjust the dust filter (52) until the average dust concentration in the observation chamber (33) is 0; Load powder into the fluidized bed (21), turn on the annular light source (34), start the air pump (41), the fluidized bed (21) starts to generate powder, and adjust the dust uniformity through the turbulence of the mixing chamber (23) so that the uniformity of the dust reaches the required level when it reaches the observation chamber (33); Among them, the dust uniformity detection is judged by the dispersion degree of the concentration data uploaded by the dust concentration meters at the four dust concentration sampling points (331). The calculation method of the dispersion degree is as follows: ； in, The population standard deviation is 1. The population mean and The mathematical expression is as follows: ； in, The number of dust concentration data points. This represents the i-th dust concentration data value; When the dust concentration dispersion degree CV calculated by the host computer (11) > CV0, it means that the dust concentration uniformity does not meet the required level. Adjust the turbulence state and intensity of the mixing chamber (23) until the dust concentration dispersion degree CV < CV0; The host computer (11) calculates the average concentration C of the four dust concentration meters. If C > C0, the flow rate of the air pump (41) is reduced to reduce the amount of dust generated by the fluidized bed (21); if C < C0, the flow rate of the air pump (41) is increased to increase the amount of dust generated by the fluidized bed (21). Each time the flow rate of the fluidized bed (21) is adjusted, the discrete coefficient needs to be re-detected. When C=C0, the camera (31) starts to capture dust images.