A video monitoring-based dust monitoring and intelligent control system and method for dry bulk cargo wharf

Abstract

The application discloses a kind of based on video monitoring's dry bulk cargo wharf dust monitoring and intelligent control system and method, belong to environmental monitoring and dust control technical field.The system includes: video acquisition module, for obtaining wharf multi-view real-time video;Image analysis module is used to fog enhancement and region growth segmentation to video image, extracts dust feature;Multi-source data fusion module uses adaptive weighted fusion algorithm, according to real-time meteorological condition and data source confidence degree dynamically adjusts weight, with laser radar, point sensor and meteorological data are deeply fused, and global dust concentration distribution chart is generated;Intelligent decision module, according to dust distribution chart, combined with mobile target recognition obstacle avoidance and operation equipment position correlation analysis, through fuzzy PID control generates accurate spraying instruction.The application realizes the leap from point monitoring to surface perception, solves the problem that traditional method covers narrow, response lag, control extensive, with significant environmental and economic benefits.

Description

Technical Field

[0001] This invention relates to the field of industrial environmental pollution prevention and control technology, and in particular to a video-based dust monitoring and intelligent control system and method for dry bulk cargo terminals. Background Technology

[0002] Dry bulk cargo terminals, as distribution centers for bulk commodities such as coal and ore, generate large amounts of fugitive dust during various stages of storage, unloading, and loading, posing a serious threat to the surrounding atmospheric environment, worker health, and equipment safety. To control dust pollution, ports are typically equipped with dust monitoring equipment and dust suppression spray systems.

[0003] Currently, dust monitoring in ports primarily relies on ground-based monitoring instruments, such as point-type monitors using beta-ray or light scattering methods. These devices are deployed at the quay boundary or inside the storage yard to monitor dust concentration in real time. When the dust concentration exceeds a preset threshold, an alarm is usually triggered or a sprinkler system is activated to suppress dust. However, these traditional monitoring methods have significant limitations in practical applications. First, their monitoring range is extremely limited; a single sensor can only reflect the concentration within a few meters of its probe. For large storage yards, achieving full coverage monitoring would require hundreds or even thousands of monitoring points, which is costly and difficult to maintain, resulting in sparse deployment and numerous blind spots. Second, point-type monitoring suffers from response lag; it can only be detected and triggered when dust spreads and reaches the sensor probe, failing to provide early warning at the moment dust rises, thus missing the optimal time for dust suppression. Furthermore, existing monitoring methods are insufficient to trace and visualize the dust diffusion path. Managers cannot intuitively understand the dust generation source, diffusion direction, and distribution range, resulting in subsequent dust suppression measures lacking specificity and often becoming "post-event remediation" rather than "pre-event prevention."

[0004] To address these issues, several technological explorations have been undertaken in related fields. For example, some research has attempted to introduce video surveillance systems at docks, transmitting live images from closed-circuit television cameras to a central control room for manual visual observation to determine dust levels. While this method provides intuitive visual information, it heavily relies on human subjective experience and attention, and is susceptible to changes in lighting and weather conditions (such as rain, fog, and haze), making it difficult to achieve all-weather, objective, and quantitative monitoring. Furthermore, some technical solutions propose using image recognition technology to automatically determine dust levels; however, these solutions typically fail to consider image correction under complex weather conditions, leading to a sharp decline in recognition accuracy in adverse weather. More importantly, existing technologies lack effective deep fusion between image recognition results and physical sensor data; the two are often used independently or simply superimposed, resulting in insufficient accuracy in dust concentration retrieval and inadequate overall system reliability. Meanwhile, the traditional sprinkler dust suppression systems that are used in conjunction with these systems mostly adopt timed control or simple logic control based on single-point thresholds. They have a low level of intelligence and often result in phenomena such as "dust has started but the sprinklers have not been turned on" or "dust-free areas are being sprayed without water." This not only causes a huge waste of water resources, but may also lead to secondary pollution of the dock surface and equipment safety hazards due to water accumulation.

[0005] Therefore, there is an urgent need in this field for a highly efficient dust monitoring technology that can achieve large-area, real-time, and accurate monitoring and can be intelligently linked with dust suppression equipment to make up for the shortcomings of existing technologies. Summary of the Invention

[0006] This invention aims to overcome the shortcomings of existing technologies and provide a dust monitoring and control system for dry bulk cargo terminals that features a wide monitoring range, fast response speed, high measurement accuracy, and intelligent control. Through non-contact video monitoring and multi-source information fusion technology, this invention achieves "visualization, knowability, and controllability" of dust emissions throughout the entire terminal operation area, providing technical support for the refined management of the port environment.

[0007] To solve the above-mentioned technical problems, the technical solution proposed in this application is as follows:

[0008] This invention provides a video-based dust monitoring and intelligent control system for dry bulk cargo terminals, comprising: The video acquisition module is used to acquire real-time video data of the dock area; The image analysis module is used to process the video data and extract dust features; A multi-source data fusion module is used to integrate the analysis results of the image analysis module with data from other sensors to generate a dust distribution map; The intelligent decision-making module is used to generate control instructions for the dust suppression equipment based on the dust distribution map.

[0009] Furthermore, the video acquisition module includes multiple high-definition network cameras deployed at high points in the dock area, which are used to acquire real-time video streams from multiple perspectives.

[0010] Furthermore, the image analysis module is specifically used to perform image dehazing based on dark channel prior and dust region segmentation based on region growing. When performing image dehazing based on dark channel prior, image details obscured by atmospheric scattering are restored by estimating atmospheric light values ​​and non-uniform transmittance. When performing dust region segmentation based on region growing, a region growing algorithm combining texture complexity and gray-level statistical features is adopted, using bright and uniformly textured regions in the image as seed points, and performing region growing based on gray-level similarity and spatial proximity between pixels.

[0011] Furthermore, the multi-source data fusion module employs an adaptive weighted fusion algorithm to assign dynamic weights to image inversion concentration, lidar data, and point sensor data. These weights are dynamically adjusted based on meteorological conditions and the real-time confidence levels of each data source. The multi-source data fusion module is also used to access real-time wind speed and direction data provided by meteorological stations and to predict dust diffusion trends based on this data.

[0012] Furthermore, the intelligent decision-making module integrates moving target recognition and obstacle avoidance functions. It identifies workers or vehicles within the grid through a real-time target detection algorithm and avoids these moving targets when generating spraying instructions.

[0013] Furthermore, the intelligent decision-making module is also used to interact with the dock production management system to obtain the geographical location of the operating machinery in real time, and to perform spatial correlation analysis between the dust distribution map and the geographical location of the operating machinery in order to locate the mobile dust source.

[0014] Furthermore, the intelligent decision-making module includes a fuzzy PID controller, which fuzzily infers the spraying frequency and water volume based on the degree of dust concentration exceeding the standard and its changing trend.

[0015] On the other hand, this application also claims protection for a video-based dust monitoring and intelligent control method for dry bulk cargo terminals, which includes the following steps: S1: Collect multi-view video data and multi-source sensor data of the dock; S2: Perform dehazing, enhancement, and segmentation processing on the acquired video images to retrieve the preliminary dust concentration; S3: The image inversion results are fused with multi-source sensor data to generate a high-precision dust distribution heat map; S4: Based on the generated dust distribution heat map and the real-time operation status of the dock, implement precise dust suppression control.

[0016] Furthermore, the dehazing process in step S2 employs an improved dark channel prior algorithm, which restores image details obscured by atmospheric scattering by estimating atmospheric light values ​​and non-uniform transmittance; the segmentation process in step S2 employs a region growing algorithm that combines texture complexity and grayscale statistical features, using bright and uniformly textured regions in the image as seed points, and performing region growing based on grayscale similarity and spatial proximity between pixels to accurately outline the dust clump contour.

[0017] Furthermore, the fusion in step S3 adopts an adaptive weighted fusion algorithm to assign dynamic weights to image inversion concentration, lidar data, and point sensor data. The weights are dynamically adjusted according to meteorological conditions and the real-time confidence of each data source.

[0018] Furthermore, the precise dust suppression control in step S4 includes: identifying moving targets within the dust suppression area and performing obstacle avoidance; and / or, obtaining the real-time geographical location of the operating machinery, and performing spatial correlation analysis between the dust distribution heat map and the geographical location of the operating machinery to locate the moving dust source and perform priority dust suppression.

[0019] Furthermore, the precise dust suppression control in step S4 also includes: using a fuzzy PID controller to fuzzily infer the spraying frequency and water volume based on the degree of dust concentration exceeding the standard and its changing trend.

[0020] Furthermore, the multi-source sensor data in step S1 includes lidar data, point dust monitor data, and weather station data.

[0021] Compared with the prior art, the present invention achieves the following beneficial technical effects: This application achieves a leap from point-based monitoring to full-area perception, effectively eliminating monitoring blind spots; it ensures monitoring reliability in complex environments around the clock through an adaptive fusion mechanism; it upgrades the response mode from passive and lagging to proactive early warning, and realizes accurate tracing of dust sources and intelligent coordination of dust suppression resources, significantly improving governance efficiency and saving resources. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1This is a structural block diagram of a video-based dust monitoring and intelligent control system for dry bulk cargo terminals, provided as an embodiment of the present invention.

[0024] Figure 2 This is a flowchart of the video image processing of the image analysis module of the present invention.

[0025] Figure 3 This is a diagram of the multi-source data adaptive weighted fusion model of the present invention.

[0026] Figure 4 The gridded dust suppression control logic diagram of this invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] like Figure 1 As shown in the figure, this embodiment provides a dust monitoring and intelligent control system for dry bulk cargo terminals based on video monitoring. The system specifically includes a video acquisition module, an image analysis module, a multi-source data fusion module, and an intelligent decision-making module.

[0029] The video acquisition module is used to acquire real-time video data of the dock area. In this embodiment, to address the limited field of view of a single camera, multiple high-definition network cameras are strategically deployed at high points in the dock area, such as above the windbreak netting and on top of high-mast lights, to construct a fully covered monitoring network. These cameras have 360° horizontal rotation and ±90° vertical rotation capabilities and support H.265 encoding to achieve high-definition video transmission under low bandwidth. The video streams acquired by all cameras are transmitted to the central server in real time via the RTSP protocol.

[0030] The image analysis module is connected to the video acquisition module and is used to process the video data and extract dust features. This module receives the video stream from the video acquisition module and performs a series of complex image processing operations on it, including image dehazing, enhancement, and dust region segmentation. Finally, it extracts multiple quantitative features from the segmented dust regions, such as the average grayscale value of the region, the total area of ​​pixels in the region, texture energy features, and regional deformation features between adjacent frames. These features together constitute a quantitative description of the dust concentration and diffusion state.

[0031] The multi-source data fusion module is connected to the image analysis module and various sensors to integrate the analysis results from the image analysis module with data from other sensors to generate a dust distribution map. In this embodiment, in addition to receiving dust feature data from the image analysis module, the module also receives data from the lidar scanning system, data from the benchmark dust monitor, and environmental data such as wind speed, wind direction, temperature, and humidity provided by the weather station. By fusing these multi-source data, a gridded heat map of dust concentration distribution covering the entire dock area is finally output.

[0032] The intelligent decision-making module is connected to the multi-source data fusion module and is used to generate control commands for dust suppression equipment based on the dust distribution map. This module receives a dust distribution heat map in real time and divides the entire wharf area into several independently controlled grids. When the concentration in a certain grid exceeds a preset threshold, it combines the moving target identification results and the location information of the operating equipment to generate precise control commands for the gridded spraying system, thereby achieving precise delivery of dust suppression resources.

[0033] Through the coordinated work of the above four modules, this invention constructs a complete "perception-analysis-decision-execution" closed-loop control system, realizing full-domain monitoring and intelligent control of dust at dry bulk cargo terminals.

[0034] In one embodiment of this application, the video acquisition module specifically includes multiple high-definition network cameras deployed at high points in the dock area to acquire real-time video streams from multiple perspectives.

[0035] Specifically, in this embodiment, multiple high-definition network eagle-eye cameras were deployed at high points such as the top of the windbreak netting and the top of the high-mast lights. These cameras have 360° horizontal rotation and ±90° vertical rotation capabilities, enabling panoramic coverage and blind-spot-free monitoring. To generate a seamless panoramic view, the video acquisition module also runs an image stitching algorithm based on SIFT feature matching. This algorithm can spatially align and fuse video frames from multiple cameras to form a unified digital panoramic image of the entire wharf, providing a precise spatial reference for subsequent regional dust analysis.

[0036] In one embodiment of this application, the image analysis module is specifically used to perform image dehazing based on dark channel prior and dust region segmentation based on region growing. For example... Figure 2 The video image processing flowchart shown below details the processing flow of this module.

[0037] First, to address the issue of decreased image contrast caused by moisture and haze in the dock environment, this module employs an improved dark channel prior algorithm to dehaze the input raw image. This algorithm, by estimating the atmospheric light value and non-uniform transmittance of the image, effectively restores image details obscured by atmospheric scattering, significantly improving the visibility between dusty areas and the background. After processing with the dark channel prior algorithm, the haze in the image is effectively removed, and the distinction between foreground and background is more pronounced. To further enhance the contrast between dust and background while suppressing excessive background noise, this module also applies a contrast-limited adaptive histogram equalization algorithm to enhance the image after dehazing.

[0038] Next, based on image enhancement, the crucial dust region segmentation is performed. This module employs a region growing algorithm that combines texture complexity and grayscale statistical features. The algorithm first automatically identifies "seed points" in the enhanced image. These seed points are typically located in areas of high brightness and relatively uniform internal texture, which aligns with the typical visual characteristics of dust clumps. Then, starting from these seed points, the algorithm continuously merges surrounding pixels that meet the criteria into a connected region based on preset similarity criteria, including grayscale similarity between pixels, spatial proximity, and consistency of texture features. This region growing method can accurately delineate the contours of dust clumps with varied shapes and irregular edges. Compared to traditional thresholding or edge detection methods, it exhibits higher robustness and accuracy, which is beneficial for dust feature extraction and output.

[0039] In one embodiment of this application, the multi-source data fusion module adopts an adaptive weighted fusion algorithm to assign dynamic weights to image inversion concentration, lidar data, and point sensor data. The weights are dynamically adjusted according to meteorological conditions and the real-time confidence level of each data source. At the same time, the module is also used to access real-time wind speed and wind direction data provided by the meteorological station and predict the dust diffusion trend based on these data.

[0040] like Figure 3As shown, the core of this multi-source data fusion module is an adaptive weighted data fusion algorithm, designed to overcome the limitations of a single sensor in complex environments. The weights are calculated based on the following logic: Regarding adaptive illumination, the system analyzes the brightness histogram of video frames to determine the current illumination conditions. At night or under low light conditions, the confidence level of image inversion density decreases significantly, and the weight assigned to the image inversion data is correspondingly reduced, while the weight of the LiDAR data is increased. Regarding adaptive meteorological conditions, when humidity is high or there is rain or fog, the confidence level of image and LiDAR data may decrease, and the algorithm dynamically adjusts the weights based on humidity sensor data. When wind speed increases, the local representativeness of point sensors may decrease, and their weights are dynamically lowered, while the weights of image and LiDAR data, which reflect trends over a larger area, are relatively increased. Regarding data consistency verification, the algorithm compares the values ​​between different data sources in real time. If the value of a certain data source deviates significantly from other data sources, the algorithm considers its confidence level lower and reduces its weight, avoiding overall misjudgment due to a single sensor malfunction.

[0041] After the above-mentioned environmentally adaptive dynamic weighted fusion calculation, the final algorithm outputs a gridded heat map of dust concentration distribution covering the entire dock area. Simultaneously, this module, combined with real-time wind speed and direction data, uses a fluid dynamics model or particle diffusion algorithm to predict the dust diffusion trend in the near future, achieving short-term early warning.

[0042] In one embodiment of this application, the intelligent decision-making module integrates moving target recognition and obstacle avoidance functions. It identifies workers or vehicles within the grid through a real-time target detection algorithm and avoids these moving targets when generating spraying instructions.

[0043] Specifically, such as Figure 4 As shown, the intelligent decision-making module integrates a real-time target detection algorithm based on YOLOv5. Before issuing a spraying command to a grid, the module calls the video stream corresponding to that grid and uses the YOLOv5 algorithm to identify whether there are moving targets within the grid, such as on-site workers or mobile machinery. If personnel or vehicles are detected, the system automatically executes obstacle avoidance strategies: either by briefly delaying the time until the target leaves the spraying area, or by dynamically adjusting the spray angle of specific sprinkler heads within the grid to accurately avoid the target, thereby ensuring that the spraying operation will not cause adverse effects or safety hazards to personnel and equipment.

[0044] In one embodiment of this application, the intelligent decision-making module is also used to interact with the dock production management system to obtain the geographical location of the operating machinery in real time, and to perform spatial correlation analysis between the dust distribution map and the geographical location of the operating machinery in order to locate the mobile dust source.

[0045] Specifically, such as Figure 4 As shown, the control system interacts with the terminal's production management system in real time via a network interface to obtain material information for each stack in the yard and the real-time geographical coordinates of operating machinery such as reclaimers, ship loaders, and stackers. Through spatial correlation analysis, the system can overlay the current dust concentration heat map with the location of the operating machinery. If a high-concentration area coincides with the location of a reclaimer that is currently operating, the system can accurately pinpoint that equipment as the main mobile dust source. At this time, the intelligent decision-making module will prioritize and focus on intensifying spraying on the grid where the equipment is located and the downwind grid, achieving precise intervention of "dust control as soon as it arises," suppressing the generation and spread of dust at its source.

[0046] In one embodiment of this application, the intelligent decision-making module includes a fuzzy PID controller, which fuzzy infers the spraying frequency and water volume based on the degree of dust concentration exceeding the standard and its changing trend.

[0047] Specifically, such as Figure 4 As shown, after deciding to start the spraying, the specific execution parameters are generated by a fuzzy PID controller. This controller takes the current dust concentration exceeding the standard (error E) and the concentration change trend (error change rate EC) as input. Based on a preset fuzzy control rule base, such as "if the concentration is very high and rises extremely rapidly, output the maximum spray water volume and the highest spray frequency," the optimal control quantities are fuzzily inferred, namely the opening frequency and single opening duration of the spray solenoid valve. This control method allows the dust suppression system to dynamically adjust the "impact intensity" according to the severity and development trend of the dust, thereby maximizing water conservation while ensuring dust suppression effectiveness.

[0048] Based on the aforementioned video-based dust monitoring and intelligent control system for dry bulk cargo terminals, this application also provides a video-based dust monitoring and intelligent control method for dry bulk cargo terminals, including the following steps S1 to S4.

[0049] Step S1: Collect multi-view video data and multi-source sensor data of the dock. In this step, multi-view video data is collected by multiple high-definition network cameras deployed at the highest point of the dock, and multi-source sensor data is collected by equipment such as lidar, point dust monitors, and weather stations.

[0050] Step S2: Perform dehazing, enhancement, and segmentation processing on the acquired video images to invert the initial dust concentration. In this step, the image analysis module performs dehazing processing based on dark channel prior, enhancement processing based on CLAHE, and dust region segmentation based on region growing on the video images. It then extracts grayscale, texture, and other features from the segmented dust regions to invert the initial dust concentration.

[0051] Step S3: The image inversion results are fused with multi-source sensor data to generate a high-precision dust distribution heat map. In this step, the multi-source data fusion module uses an adaptive weighted fusion algorithm to dynamically weight and fuse image inversion concentration, lidar data, and point sensor data, and combines them with meteorological data to generate a gridded dust concentration distribution heat map covering the entire wharf.

[0052] Step S4: Based on the generated dust distribution heat map and the real-time operational status of the dock, execute precise dust suppression control. In this step, the intelligent decision-making module generates precise control commands for the gridded spray system based on the dust distribution heat map, combined with the moving target recognition results and the real-time location of the operating equipment, and executes the dust suppression operation.

[0053] Therefore, the steps of the video-based dust monitoring and intelligent control method for dry bulk cargo terminals correspond one-to-one with the aforementioned video-based dust monitoring and intelligent control system modules for dry bulk cargo terminals. The method flow depends on the coordinated operation of the system hardware and software modules. The two belong to the same inventive concept and together constitute the complete technical solution of this invention.

[0054] In one embodiment of this application, the dehazing process in step S2 of this embodiment adopts an improved dark channel prior algorithm, which restores the image details obscured by atmospheric scattering by estimating atmospheric light value and non-uniform transmittance; the segmentation process in step S2 adopts a region growing algorithm that combines texture complexity and gray-level statistical features, using bright and uniform textured regions in the image as seed points, and performing region growing based on gray-level similarity and spatial proximity between pixels to accurately outline the dust clump contour.

[0055] Regarding the specific implementation of dehazing: The dark channel prior algorithm, based on statistical observations of a large number of fog-free outdoor images, found that in most non-sky local regions, at least one color channel has a very low intensity value, approaching zero. In foggy images, due to the scattering effect of atmospheric light, the intensity values ​​of these dark channels increase. By estimating atmospheric light values ​​and non-uniform transmittance, fog-free images can be recovered from foggy images. The improved algorithm used in this embodiment has optimized parameters for the unique water vapor and dust mixing environment of the dock, and can more effectively restore image details obscured by both sea fog and dust.

[0056] Regarding the specific implementation of the segmentation process: The region growing algorithm first automatically identifies seed points in the enhanced image. These seed points satisfy two conditions: "high brightness" and "low texture complexity," which are consistent with the typical visual characteristics of dust clumps. Then, starting from the seed points, the algorithm grows according to the 8-neighborhood connectivity. The growth criteria include: the gray-level difference between the pixel to be grown and the seed region is less than a preset threshold, and the local texture complexity of the pixel to be grown is lower than a preset threshold. In this way, the algorithm can accurately delineate the contours of dust clumps with varied shapes and irregular edges, avoiding the oversegmentation or undersegmentation problems that are prone to occur in traditional segmentation methods.

[0057] In one embodiment of this application, the fusion in step S3 adopts an adaptive weighted fusion algorithm to assign dynamic weights to image inversion concentration, lidar data, and point sensor data. The weights are dynamically adjusted according to meteorological conditions and the real-time confidence of each data source.

[0058] Regarding the specific implementation of adaptive weighted fusion: In this embodiment, the fusion algorithm uses the following formula to calculate the final concentration value of each grid cell: C final = w img C img + w lidar C lidar + w point C point ,in w img + w lidar + w point =1. The weight is determined based on real-time environmental parameters and sensor status: when the light intensity I is below the threshold L. th hour, w img Decrease linearly, w lidar The relative humidity H increases accordingly; the relative humidity H is higher than the threshold H. th hour, w img and w lidar Decrease w point Increase; wind speed W exceeds threshold W th hour, w point Decrease w imgand w lidar Increase; when the deviation of a sensor's reading from the average value of other sensors exceeds a threshold D. th When this happens, the sensor's weight is temporarily set to zero until the data returns to normal. Through this dynamic adjustment mechanism, the system can always select the most reliable combination of data sources under the current environment, ensuring the accuracy and stability of concentration inversion.

[0059] In one embodiment of this application, the precise dust suppression control in step S4 includes: identifying moving targets within the dust suppression area and performing obstacle avoidance; and / or, obtaining the real-time geographical location of the operating machinery, and performing spatial correlation analysis between the dust distribution heat map and the geographical location of the operating machinery to lock the moving dust source and perform priority dust suppression.

[0060] Regarding moving target identification and obstacle avoidance: As mentioned above, this embodiment uses the YOLOv5 algorithm to identify workers or vehicles within the sprinkler grid in real time. When a moving target is detected, the system executes one of two obstacle avoidance strategies: if the target is passing by briefly, the spraying is delayed by N seconds until the target leaves; if the target remains for a long time, the spraying angles of multiple sets of sprinkler heads within the grid are adjusted to form a spraying pattern that avoids the target, ensuring that the sprinkler operation does not affect the safety of personnel and equipment.

[0061] Regarding mobile dust source locking and priority dust suppression: This embodiment integrates with the dock production management system to obtain the GPS coordinates of each operating machine in real time. When the system detects that the dust concentration in a certain grid exceeds the standard, and the coordinates of that grid coincide with the coordinates of a material reclaimer that is currently operating, the system marks that material reclaimer as the "current primary dust source." At this time, the system not only sprays that grid but also automatically activates the spraying equipment of 1-2 adjacent grids downwind of that grid, forming a dual protection of "dust source suppression + diffusion interception." Simultaneously, the system increases the sampling frequency and control priority of that dust source to ensure that dust generation is controlled at the source.

[0062] In one embodiment of this application, the precise dust suppression control in step S4 of this embodiment further includes: using a fuzzy PID controller to fuzzily infer the spraying frequency and water volume based on the degree of dust concentration exceeding the standard and its changing trend.

[0063] Regarding the specific implementation of fuzzy PID control: In this embodiment, the input variables of the fuzzy PID controller are the concentration deviation E (the difference between the current concentration and the threshold) and the deviation change rate EC (the change in concentration per unit time). The output variables are the spraying frequency F and the spraying duration T. The fuzzy control rule base is established based on expert experience. For example, if E is positive and EC is positive, then F is maximum and T is maximum; if E is positive and EC is negative, then F is moderate and T is moderate. Through the processes of fuzzification, fuzzy inference, and defuzzification, the controller outputs precise F and T values, driving the spraying equipment to operate in an optimal manner. Actual measurements show that compared to traditional on / off control, fuzzy PID control can save more than 30% of water while increasing the dust concentration compliance rate to over 95%.

[0064] In one embodiment of this application, the multi-source sensor data in step S1 includes lidar data, point dust monitor data, and weather station data.

[0065] Specifically, this embodiment deploys the following sensors: a lidar scanning system, deployed at high points around the yard, which obtains average concentration information along a line by analyzing the backscattering extinction coefficient of laser light on dust particles; point-type dust monitors, using the beta-ray method or light scattering method, deployed at the dock boundary and key operating points as a "true value" reference for concentration calibration; and a weather station, deployed in an open area, providing real-time environmental parameters such as wind speed, wind direction, temperature, humidity, and atmospheric pressure. These three types of data, together with video image data, form the basis of multi-source data fusion, providing data support for generating high-precision, high-reliability dust distribution maps.

[0066] In summary, the present invention provides a video-based dust monitoring and intelligent control system and method for dry bulk cargo terminals. Through adaptive weighted fusion of video, radar, point sensors, and meteorological data, it achieves an upgraded perception of terminal dust from a "point" to a "surface" perspective and strong adaptability to complex environments. By integrating dust source tracking with the production management system and combining it with visual recognition for intelligent obstacle avoidance, it achieves a leap from passive response to proactive source tracing and precise control. The overall combination of these technical features constitutes a significant advancement over existing technologies, solving a long-standing technical challenge in the specific scenario of dry bulk cargo terminals, and possesses both non-obviousness and significant inventiveness.

[0067] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A video-based dust monitoring and intelligent control system for dry bulk cargo terminals, characterized in that, It includes: The video acquisition module is used to acquire real-time video data of the dock area; The image analysis module is used to process the video data and extract dust features; A multi-source data fusion module is used to integrate the analysis results of the image analysis module with data from other sensors to generate a dust distribution map; The intelligent decision-making module is used to generate control instructions for the dust suppression equipment based on the dust distribution map.

2. The video-based dust monitoring and intelligent control system for dry bulk cargo terminals according to claim 1, characterized in that, The video acquisition module includes multiple high-definition network cameras deployed at high points in the dock area. These high-definition network cameras are used to acquire real-time video streams from multiple perspectives.

3. The dust monitoring and intelligent control system for dry bulk cargo terminals based on video monitoring according to claim 1, characterized in that, The image analysis module is specifically used to perform image dehazing based on dark channel prior and dust region segmentation based on region growing. Specifically, when performing image dehazing based on dark channel prior, image details obscured by atmospheric scattering are restored by estimating atmospheric light values ​​and non-uniform transmittance. When performing dust region segmentation based on region growing, a region growing algorithm combining texture complexity and gray-level statistical features is adopted. High-brightness and uniformly textured regions in the image are used as seed points, and region growing is performed based on the gray-level similarity and spatial proximity between pixels.

4. The dust monitoring and intelligent control system for dry bulk cargo terminals based on video monitoring according to claim 1, characterized in that, The multi-source data fusion module employs an adaptive weighted fusion algorithm to assign dynamic weights to image inversion concentration, lidar data, and point sensor data. These weights are dynamically adjusted based on meteorological conditions and the real-time confidence levels of each data source. The multi-source data fusion module is also used to access real-time wind speed and direction data provided by meteorological stations and to predict dust diffusion trends based on this data.

5. The dust monitoring and intelligent control system for dry bulk cargo terminals based on video monitoring according to claim 1, characterized in that, The intelligent decision-making module integrates moving target recognition and obstacle avoidance functions. It identifies workers or vehicles within the grid through a real-time target detection algorithm and avoids these moving targets when generating spraying instructions.

6. The video-based dust monitoring and intelligent control system for dry bulk cargo terminals according to claim 1, characterized in that, The intelligent decision-making module is also used to interact with the dock production management system to obtain the geographical location of the operating machinery in real time, and to perform spatial correlation analysis between the dust distribution map and the geographical location of the operating machinery in order to locate the mobile dust source.

7. The video-based dust monitoring and intelligent control system for dry bulk cargo terminals according to claim 5 or 6, characterized in that, The intelligent decision-making module includes a fuzzy PID controller, which fuzzily infers the spraying frequency and water volume based on the degree of dust concentration exceeding the standard and its changing trend.

8. A method for dust monitoring and intelligent control at dry bulk cargo terminals based on video surveillance, characterized in that, It includes the following steps: S1: Collect multi-view video data and multi-source sensor data of the dock; S2: Perform dehazing, enhancement, and segmentation processing on the acquired video images to retrieve the preliminary dust concentration; S3: The image inversion results are fused with multi-source sensor data to generate a high-precision dust distribution heat map; S4: Based on the generated dust distribution heat map and the real-time operation status of the dock, implement precise dust suppression control.

9. The method for dust monitoring and intelligent control at dry bulk cargo terminals based on video monitoring according to claim 8, characterized in that, The dehazing process in step S2 employs an improved dark channel prior algorithm, which restores image details obscured by atmospheric scattering by estimating atmospheric light values ​​and non-uniform transmittance. The segmentation process in step S2 employs a region growing algorithm that combines texture complexity and grayscale statistical features. Using bright and uniformly textured regions in the image as seed points, region growing is performed based on the grayscale similarity and spatial proximity between pixels to accurately outline the dust clumps.

10. The method for dust monitoring and intelligent control at dry bulk cargo terminals based on video monitoring according to claim 8, characterized in that, The fusion in step S3 adopts an adaptive weighted fusion algorithm, which assigns dynamic weights to image inversion concentration, lidar data, and point sensor data. The weights are dynamically adjusted according to meteorological conditions and the real-time confidence of each data source.

11. The method for dust monitoring and intelligent control at dry bulk cargo terminals based on video monitoring according to claim 8, characterized in that, The precise dust suppression control in step S4 includes: identifying moving targets within the dust suppression area and performing obstacle avoidance; and / or, obtaining the real-time geographical location of the operating machinery, and performing spatial correlation analysis between the dust distribution heat map and the geographical location of the operating machinery to locate the moving dust source and perform priority dust suppression.

12. The method for dust monitoring and intelligent control at dry bulk cargo terminals based on video monitoring according to claim 11, characterized in that, The precise dust suppression control in step S4 also includes: using a fuzzy PID controller to fuzzily infer the spraying frequency and water volume based on the degree of dust concentration exceeding the standard and its changing trend.

13. The method for dust monitoring and intelligent control at dry bulk cargo terminals based on video monitoring according to claim 8, characterized in that, The multi-source sensor data in step S1 includes lidar data, point dust monitor data, and weather station data.

Images