A dust control system for storage yard

Abstract

The present invention discloses a dust control system for a storage yard, including a camera module, an execution module, a splicing module, a noise filtering module, a background extraction module, a contour recognition module, a new dust positioning module, a diffuse dust positioning module, and a control module; the system periodically obtains a plane image in the storage yard, and performs noise filtering to obtain a real plane image; then identifies the contours of non-dust objects in the storage yard based on the background plane image; performs a differential operation on the real plane image to obtain the new dust positioning information at each sampling moment in the storage yard; obtains the diffuse dust positioning information in the storage yard based on the grayscale gradient of the real plane image; controls the operation of the spraying device to complete the dust control of the storage yard. The dust control of the storage yard is completed effectively and quickly. After adopting the control method, there is no need to densely arrange a large number of dust concentration monitors in the storage yard, which reduces the investment cost and maintenance cost, and also has little impact on the movement of vehicles and personnel operations in the storage yard, operations and other places.

Description

Technical Field

[0001] The invention relates to the technical field of dust control, in particular to a storage yard dust control system. Background Art

[0002] Dust is generated during raw material transportation, yard construction, material transfer, and production operations in industries such as steel, coking, thermal power, cement, casting, brick kilns, carbon, glass, ceramics, building materials, construction, and mining. The main hazards of dust are as follows: 1) harming human health; 2) causing dust explosions; 3) reducing visibility, and dust is an important component of haze; 4) dirtying buildings; 5) corroding metal facilities; 6) increasing wear and tear of machinery and equipment; 7) reducing product quality; 8) causing poor plant growth. In 2018, non-point source pollution control technology was included in the "2018 National Advanced Pollution Control Technology Catalog (Air Pollution Control Field)". In 2019, cities in the north began to implement deep emission reduction implementation plans for industries such as steel and coking, requiring steel, coking and other enterprises to fully complete accurate monitoring and control of dust emissions.

[0003] At present, some industries and enterprises are promoting the use of dust concentration monitors for dust monitoring. One is installed for every 10,000 square meters in storage yards, operations and other places. If it is less than 10,000 square meters, it will be calculated as 10,000 square meters. This monitoring method can only detect the total dust concentration, PM2.5, PM10, temperature, humidity, atmospheric pressure, wind direction, and wind force at some local points where the dust concentration online monitor is installed. The detection accuracy of these local points is relatively high. However, it is only suitable for high-precision detection of dust at local points, and cannot achieve dust concentration detection at any location in the area. If you want to achieve grid-based precise monitoring of regional dust, you must densely arrange a large number of dust concentration monitors in the entire storage yard, operation and other places, which will have high investment costs, complex installation, and high maintenance costs. The installation bracket will seriously affect the movement of vehicles, personnel operations and other activities in storage yards, operations and other places. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a dust control system for a storage yard, which does not require densely arranged a large number of dust concentration monitors in the storage yard, thereby greatly reducing investment costs, maintenance costs, etc.

[0005] To achieve the above object, the present invention provides a dust control system for a storage yard, comprising:

[0006] A camera module, comprising a plurality of camera devices arranged on the top of the storage yard, and each spatial point in the storage yard is located within the field of view of at least one camera device;

[0007] An execution module includes a plurality of spraying devices arranged on the top of the storage yard, and each spatial point in the storage yard is located within the spraying area of ​​at least one spraying device;

[0008] A stitching module is connected to the camera module for acquiring the field of view images taken by each camera device, and stitching the field of view images at the same sampling time to obtain a sampling plane image of the storage yard;

[0009] A noise filtering module is connected to the splicing module for performing noise filtering on the sampled plane image to obtain a real plane image after the noise of the camera lens is filtered out;

[0010] A background extraction module is connected to the noise filtering module to obtain a background plane image of the storage yard based on the real plane image;

[0011] A contour recognition module is connected to the background extraction module for recognizing the contours of non-dust objects in the storage yard based on the background plane image;

[0012] A new dust positioning module is added, which is connected to the noise filtering module and the contour recognition module, and is used to perform differential operations on the real plane images of all adjacent sampling moments to obtain a grayscale image sequence, and obtain the new dust positioning information at each sampling moment in the storage yard based on the grayscale image sequence;

[0013] The diffuse dust positioning module is connected to the noise filtering module to obtain the diffuse dust positioning information at each sampling moment in the storage yard based on the gray gradient of each real plane image;

[0014] The control module is connected to the new dust positioning module, the diffuse dust positioning module and the execution module to control the operation of the spraying device according to the positioning information of the new dust and diffuse dust at each sampling moment in the yard to complete the dust control in the yard.

[0015] In one embodiment, the noise filtering module includes:

[0016] A first grid division submodule is connected to the splicing module for grid division of all sampled plane images;

[0017] The grayscale mean calculation module is connected to the first grid division submodule for calculating the true grayscale mean of each grid area in the sampling plane image. The calculation process is as follows:

[0018] Δg i，j =g i，j -g j0

[0019] In the formula, Δg i，j is the true grayscale mean of the jth grid region in the i-th sampling plane image, g i，j is the grayscale mean of the jth grid region in the i-th sampling plane image, g j0 is the lowest grayscale mean of the jth grid area in all sampled plane images within a certain time period;

[0020] The grayscale mean replacement module is connected to the grid division submodule and the grayscale mean calculation module to use Δg in the sampling plane image i. i，j Replace g i，j , that is, the true plane image of the sampled plane image i is obtained.

[0021] In one embodiment, the background extraction module is embedded with the ViBe algorithm, and the background extraction process is as follows:

[0022] For any sampling moment, all real plane images within n sampling periods before the sampling moment are obtained, and the ViBe algorithm is performed on all the acquired real plane images to obtain the background plane image at the sampling moment.

[0023] In one embodiment, the process of the contour recognition module recognizing the contour of the non-dust object in the storage yard based on the background plane image is as follows:

[0024] Extracting a contour point set graphic whose grayscale value is greater than a first grayscale threshold in the background plane image;

[0025] If the contour in the contour point set graph has a certain regular geometric shape, it can be determined that this type of contour corresponds to the contour of non-dust objects in the yard, including material shed structures, operating equipment, and parked vehicles;

[0026] If the contour in the contour point set graph has an uncertain regular geometric shape, it can be determined that this type of contour corresponds to the contour of a non-dusty object including a stockpile in the stockpile.

[0027] In one embodiment, the newly added dust positioning module includes:

[0028] The point cloud extraction submodule is connected to the noise filtering module for extracting contour point set graphics whose grayscale values ​​are greater than a second grayscale threshold in the grayscale image sequence to obtain a contour point set graphics set;

[0029] The point cloud removal submodule is communicated with the contour recognition module and the point cloud extraction submodule to remove the contour point set graphics corresponding to the contour of the non-dust object from the contour point set graphics set. At this time, the remaining contour point set graphics in the contour point set graphics set correspond to the newly added dust in the yard, that is, the newly added dust positioning information at each sampling moment in the yard is obtained.

[0030] In one embodiment, the diffuse dust positioning module includes:

[0031] A second grid division submodule, which is in communication with the noise filtering module, is used to perform grid division on any real plane image i;

[0032] A first grayscale gradient calculation module is connected to the second grid division submodule for obtaining the grayscale gradient of each grid area in the real plane image i;

[0033] The first diffuse dust judgment module is communicated with the first grayscale gradient calculation module and the second grid division submodule to extract the grid area whose grayscale gradient is less than the first grayscale gradient threshold in the real plane image i, and determine that the local area in the yard corresponding to this part of the grid area has diffuse dust, that is, to obtain the first positioning information of the diffuse dust in the yard at the sampling time i.

[0034] In one embodiment, the process of obtaining the grayscale gradient of each grid area in the real plane image i in the first grayscale gradient calculation module is specifically as follows:

[0035] gg i，j =(Δg i，j -Δg i，j+1 ) / Distance

[0036] In the formula, Δg i，j , Δg i，j+1 is the real grayscale mean of the jth and j+1th grid regions in the i-th real plane image, gg i，j is the grayscale gradient value between the jth grid pattern domain and the j+1th grid pattern domain in the i-th real plane image, and Distance is the distance between the jth grid pattern domain and the j+1th grid pattern domain;

[0037] Taking the current grid (i, j) as the center point, calculate the grayscale gradient values ​​of all grids within the distance Distance, and take the statistical mean of all grayscale gradient values ​​as the grayscale gradient of grid area j.

[0038] In one embodiment, the diffuse dust positioning module includes:

[0039] The second grayscale gradient calculation module is connected to the contour recognition module for extracting all non-dust object contours in the real plane image i and obtaining the grayscale gradient of the contour edge of each non-dust object;

[0040] The second diffuse dust judgment module is communicated with the second grayscale gradient calculation module and the second grid division submodule to extract the contours of non-dust objects whose grayscale gradient is less than the second grayscale gradient threshold, and determine that the local area in the yard corresponding to the contours of these non-dust objects has diffuse dust, so as to obtain the second positioning information of the diffuse dust in the yard at the sampling time i.

[0041] In one embodiment, a plurality of the spraying devices and the camera devices are arranged in a grid-like manner on top of the storage yard.

[0042] In one embodiment, the camera device includes at least one of a gun camera, a ball camera, a gun-ball camera, and a panoramic camera.

[0043] The present invention provides a dust control system for a storage yard, which periodically obtains a plane image of the storage yard, and obtains positioning information of diffuse dust and new dust in the storage yard in a two-dimensional plane based on image recognition and processing of the plane image in the storage yard, and controls the operation of a spraying device based on the positioning information, thereby effectively and quickly completing the dust control in the storage yard. By adopting the control system, there is no need to densely arrange a large number of dust concentration monitors in the storage yard, which reduces investment costs and maintenance costs, and also has little impact on activities such as vehicle movement and personnel operations in storage yards, work sites, and the like. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0045] Figure 1 It is a module schematic diagram of the yard dust control system in an embodiment of the present invention.

[0046] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] It should be noted that all directional indications in the embodiments of the present invention (such as up, down, left, right, front, back, etc.) are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0049] In addition, in the present invention, descriptions such as "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0050] In the present invention, unless otherwise clearly specified and limited, the terms "connection", "fixation", etc. should be understood in a broad sense. For example, "fixation" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, an electrical connection, a physical connection, or a wireless communication connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0051] In addition, the technical solutions between the various embodiments of the present invention can be combined with each other, but it must be based on the fact that ordinary technicians in the field can implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0052] This embodiment discloses a dust control system for a storage yard, wherein a plurality of spraying devices and camera devices are arranged on the top of the storage yard as camera modules and execution modules, wherein each spatial point in the storage yard is located within the spraying area of ​​at least one spraying device and within the field of view of at least one camera device, so each point in the two-dimensional plane from top to bottom in the storage yard is located within the field of view of at least one camera device, and a plane image in the two-dimensional plane from top to bottom in the storage yard is obtained by splicing the field of view images taken by each camera device. It periodically obtains the plane image of the storage yard through the camera module, and obtains the positioning information of the diffuse dust and the newly added dust in the storage yard in the two-dimensional plane based on the image recognition and processing of the plane image in the storage yard, and controls the operation of the execution module based on this positioning information, that is, it can effectively and quickly complete the dust control of the storage yard. After adopting this control system, there is no need to densely arrange a large number of dust concentration monitors in the storage yard, which reduces the investment cost and maintenance cost, and at the same time has little impact on the vehicle movement, personnel operation and other activities in the storage yard, operation and other places.

[0053] In this embodiment, the camera module can be a grid layout of different types of cameras such as gun cameras, ball cameras, gun-ball cameras, and panoramic cameras on the top of the yard. The yard is constructed into a two-dimensional plane space through the field of view images taken by all cameras, so that it can reflect the shape, size and coordinates of objects in the yard. It is a conventional technical means to stitch multiple field of view images into a plane image, so it will not be repeated in this embodiment. For details, please refer to patents CN103597810A, CN110969594A, etc.

[0054] refer to Figure 1 The dust control system of the storage yard in this embodiment includes a camera module, an execution module, a splicing module, a noise filtering module, a background extraction module, a contour recognition module, a new dust positioning module, a diffuse dust positioning module and a control module.

[0055] Among them, the stitching module is connected to the camera module in communication to obtain the field of view images taken by each camera device, and stitches the field of view images at the same sampling time to obtain the sampling plane image of the yard; the noise filtering module is connected to the stitching module in communication to filter the sampling plane image to obtain the real plane image after filtering out the noise of the camera lens; the background extraction module is connected to the noise filtering module in communication to obtain the background plane image of the yard based on the real plane image; the contour recognition module is connected to the background extraction module in communication to recognize the contours of non-dust objects in the yard based on the background plane image; the newly added dust positioning module is connected to the noise filtering module and the contour recognition module. The modules are communicatively connected to perform differential operations on the real plane images at all adjacent sampling moments to obtain a grayscale image sequence, and obtain the location information of newly added dust at each sampling moment in the stockpile based on the grayscale image sequence and the contour of the non-dust object; the diffuse dust location module is communicatively connected to the noise filtering module to obtain the location information of diffuse dust at each sampling moment in the stockpile based on the grayscale gradient of each real plane image; the control module is communicatively connected to the newly added dust location module, the diffuse dust location module, and the execution module to control the operation of the spraying device according to the location information of newly added dust and diffuse dust at each sampling moment in the stockpile to complete the dust control in the stockpile.

[0056] In this embodiment, the noise filtering module includes a first grid division submodule, a grayscale mean calculation module and a grayscale mean replacement module. Among them, the first grid division submodule is connected to the splicing module for grid division of all sampled plane images; the grayscale mean calculation module is connected to the first grid division submodule for calculating the true grayscale mean of each grid area in the sampled plane image; the grayscale mean replacement module is connected to the grid division submodule and the grayscale mean calculation module for replacing the original grayscale mean with the true grayscale mean in the sampled plane image i, that is, obtaining the true plane image of the sampled plane image i.

[0057] In this embodiment, the background extraction module is embedded with the ViBe algorithm to complete the background image extraction.

[0058] In this embodiment, the newly added dust positioning module includes a point cloud extraction submodule and a point cloud removal submodule, wherein the point cloud extraction submodule is communicated with the noise filtering module to extract contour point set graphics whose grayscale values ​​are greater than the second grayscale threshold in the grayscale image sequence to obtain a contour point set graphics set; the point cloud removal submodule is communicated with the contour recognition module and the point cloud extraction submodule to remove the contour point set graphics corresponding to the contour of the non-dust object from the contour point set graphics set. At this time, the remaining contour point set graphics in the contour point set graphics set correspond to the newly added dust in the yard, that is, the newly added dust positioning information at each sampling moment in the yard is obtained.

[0059] In this embodiment, the diffuse dust positioning module includes a second grid division submodule, a first grayscale gradient calculation module, a first diffuse dust judgment module, a second grayscale gradient calculation module and a second diffuse dust judgment module. Among them, the second grid division submodule is connected in communication with the noise filtering module to perform grid division on any real plane image i; the first grayscale gradient calculation module is connected in communication with the second grid division submodule to obtain the grayscale gradient of each grid area in the real plane image i; the first diffuse dust judgment module is connected in communication with the first grayscale gradient calculation module and the second grid division submodule to extract the grid area in the real plane image i whose grayscale gradient is less than the first grayscale gradient threshold, and determine that the local area in the yard corresponding to this part of the grid area has diffuse dust, that is, the number of grid areas in the yard at the sampling time i is obtained. The first positioning information of diffuse dust; the second grayscale gradient calculation module is communicated with the contour recognition module to extract all non-dust object contours in the real plane image i, and obtain the grayscale gradient of the edge of each non-dust object contour; the second diffuse dust judgment module is communicated with the second grayscale gradient calculation module and the second grid division submodule to extract the non-dust object contour whose grayscale gradient is less than the second grayscale gradient threshold, and determine that the local area in the yard corresponding to the non-dust object contour has diffuse dust, and obtain the second positioning information of diffuse dust in the yard at sampling time i.

[0060] The working process of the above-mentioned yard dust control system is as follows:

[0061] Step 1: The camera module periodically captures the field of view image in the yard and sends it to the stitching module, thereby obtaining all the internal plane images of the yard within the sampling period as sampling plane images; and filtering all the sampling plane images through the noise filtering module to obtain a real plane image after filtering out the noise of the camera lens. Among them, the selection of the image sampling period depends on the speed of the video image processing hardware and software. The sampling period in this embodiment is 1 second to 5 minutes. During the sampling period, the field of view images taken by each camera at the same sampling time are stitched into a sampling plane image, and all the sampling plane images within the sampling period are combined into a sampling plane image sequence in the order of the sampling times.

[0062] During the shooting process of the camera device, dust or foreign matter on the lens may cause noise in the field of view image, which will cause noise errors in the subsequent image recognition process. Therefore, noise filtering is required to filter out the lens noise of the camera device. In this embodiment, the noise filtering process is specifically as follows:

[0063] Firstly, all sampled plane images within a certain time length are grid-divided by the first grid-dividing submodule;

[0064] Then the grayscale mean calculation module is used to calculate the true grayscale mean of each grid area in the sampling plane image. The calculation process is as follows:

[0065] Δg i，j =g i，j -g j0

[0066] In the formula, Δg i，j is the true grayscale mean of the jth grid region in the i-th sampling plane image, g i，j is the grayscale mean of the jth grid region in the i-th sampling plane image, g j0 is the lowest grayscale mean of the jth grid area in all plane images within a certain time period;

[0067] Finally, the grayscale mean replacement module is used to replace the sampled plane image i with Δg i，j Replace g i，j , that is, the true plane image of plane image i is obtained.

[0068] It should be noted that the "certain time" in the above noise filtering process can be either 1 sampling cycle or 10-1000 sampling cycles. Since the change of dust or foreign matter on the camera lens is a long process, the present embodiment preferably takes the time of 1000 sampling cycles as the above g. j0 The "certain length of time" in the acquisition process. For example, if the sampling period is 1 minute, then during the noise filtering process within 1000 minutes, gj0 are the same value.

[0069] Step 2: A background plane image of the storage yard is obtained based on the real plane image through the background extraction module, and the contour recognition module is used to recognize the contours of non-dust objects in the storage yard and the positioning information of non-dust objects based on the background plane image.

[0070] Among them, since the background extraction module is embedded with the ViBe algorithm, its background extraction process is:

[0071] For any sampling moment, all real plane images within n sampling periods before the sampling moment are obtained, and the ViBe algorithm is performed on all the acquired real plane images to obtain the background plane image at the sampling moment.

[0072] There are two main categories of non-dust objects in the stockpile. The first category is objects with regular geometric shapes, such as material shed structures, operating equipment, parked vehicles, and staff. This type of object can be directly distinguished based on the shape characteristics and iconic features of the object; the second category is objects with irregular geometric shapes, such as material piles and waste slag piles. This type of object cannot be directly distinguished based on the shape characteristics and iconic features of the object. However, these two types of objects have one thing in common, that is, the grayscale values ​​of their contours on the image are large. Based on this, this embodiment proposes a method for recognizing the contours of non-dust objects in the stockpile, that is, the process of the contour recognition module recognizing the contours of non-dust objects in the stockpile based on the background plane image is as follows:

[0073] First, extract the contour point set graphics whose grayscale values ​​are greater than the first grayscale threshold in the background plane image, so that the contour point set graphics only have the contours of the above two types of non-dust objects, and then identify each contour separately:

[0074] If the contour in the contour point set graph has a certain regular geometric shape, it can be determined that this type of contour corresponds to the contour of non-dust objects in the yard, including material shed structures, operating equipment, parked vehicles, and staff;

[0075] If the contour in the contour point set graph has an uncertain regular geometric shape, it can be determined that this type of contour corresponds to the contour of non-dust objects in the stockpile, including material piles and waste slag piles;

[0076] After identifying the contours of all non-dust objects, the positioning information of all non-dust objects in the real plane image can be obtained by combining the image features of non-dust objects such as color distribution, contour continuity, geometric association, pixel density, point cloud distribution, etc.

[0077] Step 3: The new dust positioning module is used to perform differential operations on the real plane images of all adjacent sampling moments to obtain a grayscale image sequence, and based on the grayscale image sequence and the contours of non-dust objects, the new dust positioning information at each sampling moment in the yard is obtained.

[0078] Since the positioning information can be directly obtained in the plane image, and the source of dust in the yard can only be the operating equipment, working vehicles, staff, material piles or waste slag piles, that is, the contours of non-dust objects in the above step 2, the dust can be associated with the contours of non-dust objects to facilitate the control of the subsequent spraying device. For example, if the dust in an area is caused by the staff, the dust in this area will not be sprayed, and will be sprayed after the staff leave. Based on this, this embodiment proposes a method for locating newly added dust, and the specific process is as follows:

[0079] First, the point cloud extraction submodule is used to extract the contour point set graphics whose grayscale value is greater than the second grayscale threshold in the grayscale image sequence, and obtain the contour point set graphics set, wherein the second grayscale threshold is less than the first grayscale threshold. The extraction of the contour point set graphics in this step is to filter out the background graphics in the yard;

[0080] Then, the contour point set graphics corresponding to the contours of non-dust objects are removed from the contour point set graphics set through the point cloud removal sub-module. At this time, the remaining contour point set graphics in the contour point set graphics set correspond to the newly added dust in the yard, that is, the positioning information of the newly added dust in the yard at each sampling moment in the real plane image can be obtained.

[0081] Step 4: The diffuse dust positioning module is used to obtain the diffuse dust positioning information at each sampling moment in the storage yard based on the grayscale gradient of each real plane image.

[0082] In this embodiment, the diffuse dust positioning information is divided into two categories. The first category is diffuse dust in the real image where there is no area where non-dust objects are located. The specific acquisition process of the first category of diffuse dust positioning information is as follows:

[0083] The second grid division submodule is used to grid any real plane image i, and the first grayscale gradient calculation module is used to obtain the grayscale gradient of each grid area in the real plane image i, specifically:

[0084] gg i，j =(Δg i，j -Δg i，j+1 ) / Distance

[0085] In the formula, Δg i，j , Δg i，j+1 is the real grayscale mean of the jth and j+1th grid regions in the i-th real plane image, gg i，jis the grayscale gradient value between the jth grid pattern domain and the j+1th grid pattern domain in the i-th real plane image, and Distance is the distance between the jth grid pattern domain and the j+1th grid pattern domain;

[0086] Taking the current grid (i, j) as the center point, calculate the grayscale gradient values ​​of all grids within the distance Distance, and take the statistical mean of all grayscale gradient values ​​as the grayscale gradient of grid area j;

[0087] Finally, the first diffuse dust judgment module is used to extract the grid area in the real plane image i whose grayscale gradient is less than the first grayscale gradient threshold, and it is determined that the local area in the yard corresponding to this part of the grid area has diffuse dust, that is, the first positioning information of the diffuse dust in the yard at the sampling time i is obtained.

[0088] The second category is diffuse dust in the real image where there are non-dust objects. Since this type of diffuse dust is located directly above the non-dust objects, and the grayscale value of the non-dust objects is larger, its grayscale value is greater than the grayscale value of the diffuse dust; therefore, the above-mentioned first-category regional gradient recognition method is not applicable to this type of diffuse dust. Therefore, for the positioning information of the second type of diffuse dust, this embodiment proposes a method based on contour gradient judgment, and its specific acquisition process is:

[0089] For the real plane image i, all the non-dust object contours in the real plane image i are extracted, and the grayscale gradient of the contour edge of each non-dust object is obtained by the second grayscale gradient calculation module. The calculation process of the grayscale gradient is the same as the calculation process of the first category. It only needs to discretize the contour and use a discrete point as a grid area.

[0090] The second diffuse dust judgment module extracts the contours of non-dust objects whose grayscale gradient is less than the second grayscale gradient threshold, and determines that the local area in the yard corresponding to the non-dust object contours has diffuse dust, thereby obtaining the second positioning information of the diffuse dust in the yard at sampling time i.

[0091] Step 5: The control module controls the spraying device to spray the area with new dust and diffuse dust according to the location information of the new dust and diffuse dust at each sampling moment in the yard. After the spraying is completed, it pauses for a certain period of time to wait for the fog and dust to settle, completes the dust control of the yard, and enters the next sampling cycle. That is, the two adjacent sampling cycles in this embodiment are not continuous in time, and there is a period of waiting for the fog and dust to settle between the two. Among them, the spraying device in this embodiment includes a grid-shaped layout of nozzles and their water and gas pipelines at the top of the yard, and the nozzles can move with multiple degrees of freedom. As for the specific implementation structures of the nozzles, water and gas channels, they are all conventional technical means, so they are not repeated in this embodiment.

[0092] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. All equivalent structural changes made by using the contents of the present invention specification and drawings under the inventive concept of the present invention, or directly / indirectly applied in other related technical fields are included in the patent protection scope of the present invention.

Claims

1. A dust control system for a storage yard, characterized in that: include: A camera module, comprising a plurality of camera devices arranged on the top of the storage yard, and each spatial point in the storage yard is located within the field of view of at least one camera device; An execution module includes a plurality of spraying devices arranged on the top of the storage yard, and each spatial point in the storage yard is located within the spraying area of ​​at least one spraying device; A stitching module is connected to the camera module for acquiring the field of view images taken by each camera device, and stitching the field of view images at the same sampling time to obtain a sampling plane image of the storage yard; A noise filtering module is connected to the splicing module for performing noise filtering on the sampled plane image to obtain a real plane image after the noise of the camera lens is filtered out; A background extraction module is connected to the noise filtering module to obtain a background plane image of the storage yard based on the real plane image; A contour recognition module is connected to the background extraction module for recognizing the contours of non-dust objects in the storage yard based on the background plane image; A new dust positioning module is added, which is connected to the noise filtering module and the contour recognition module, and is used to perform differential operations on the real plane images of all adjacent sampling moments to obtain a grayscale image sequence, and obtain the new dust positioning information at each sampling moment in the storage yard based on the grayscale image sequence; The diffuse dust positioning module is connected to the noise filtering module to obtain the diffuse dust positioning information at each sampling moment in the storage yard based on the gray gradient of each real plane image; The control module is connected to the newly added dust positioning module, the diffuse dust positioning module and the execution module to control the operation of the spraying device according to the positioning information of the newly added dust and diffuse dust at each sampling moment in the yard to complete the dust control in the yard; The diffuse dust positioning module includes: The second grid division submodule is connected to the noise filtering module to be used for any real plane image i Perform mesh division; The first grayscale gradient calculation module is connected to the second grid division submodule for obtaining a real plane image. i The grayscale gradient of each grid area in ; The first diffuse dust judgment module is connected to the first gray gradient calculation module and the second grid division submodule to extract the real plane image. i The grid area with a grayscale gradient less than the first grayscale gradient threshold is selected, and it is determined that the local area in the yard corresponding to this part of the grid area has diffuse dust, that is, the dust in the yard at the sampling time is obtained. i First location information of diffuse dust; The second grayscale gradient calculation module is connected to the contour recognition module to extract the real plane image i All non-dust object contours in the image are obtained, and the grayscale gradient of the edge of each non-dust object contour is obtained; The second diffuse dust judgment module is connected to the second grayscale gradient calculation module and the second grid division submodule to extract the contours of non-dust objects whose grayscale gradient is less than the second grayscale gradient threshold, and determine that the local area in the yard corresponding to the contours of the non-dust objects has diffuse dust, and obtain the number of dust particles in the yard at the sampling time. i Diffuse dust second positioning information.

2. The dust control system for the storage yard according to claim 1 is characterized in that: The noise filtering module comprises: A first grid division submodule is connected to the splicing module for grid division of all sampled plane images; The grayscale mean calculation module is connected to the first grid division submodule for calculating the true grayscale mean of each grid area in the sampling plane image. The calculation process is as follows: D g i，j = g i，j - g j0 In the formula, Δ g i，j For the i The sampling plane image j The true grayscale mean of the grid pattern domain, g i，j For the i The sampling plane image j The grayscale mean of the grid area, g j0 is the number of all sampled plane images within a certain time period. j The lowest grayscale mean value of the grid area; The grayscale mean replacement module is connected to the grid division submodule and the grayscale mean calculation module to be used in sampling plane images. i Δ g i，j replace g i，j , that is, the sampling plane image is obtained i The real plane image.

3. The dust control system for the storage yard according to claim 1 is characterized in that: The background extraction module is embedded with the ViBe algorithm, and the background extraction process is as follows: For any sampling time, obtain the n All the real plane images within a sampling period are analyzed by ViBe algorithm to obtain the background plane image at the sampling moment.

4. The dust control system for the storage yard according to claim 1 is characterized in that: The process of the contour recognition module recognizing the contours of non-dust objects in the storage yard based on the background plane image is as follows: Extracting a contour point set graphic whose grayscale value is greater than a first grayscale threshold in the background plane image; If the contour in the contour point set graph has a certain regular geometric shape, it can be determined that this type of contour corresponds to the contour of non-dust objects in the yard, including material shed structures, operating equipment, and parked vehicles; If the contour in the contour point set graph has an uncertain regular geometric shape, it can be determined that this type of contour corresponds to the contour of a non-dusty object including a stockpile in the stockpile.

5. The dust control system for the storage yard according to claim 1 is characterized in that: The newly added dust positioning module includes: The point cloud extraction submodule is connected to the noise filtering module for extracting contour point set graphics whose grayscale values ​​are greater than a second grayscale threshold in the grayscale image sequence to obtain a contour point set graphics set; The point cloud removal submodule is communicated with the contour recognition module and the point cloud extraction submodule to remove the contour point set graphics corresponding to the contour of the non-dust object from the contour point set graphics set. At this time, the remaining contour point set graphics in the contour point set graphics set correspond to the newly added dust in the yard, that is, the newly added dust positioning information at each sampling moment in the yard is obtained.

6. The dust control system for the storage yard according to claim 1 is characterized in that: The first grayscale gradient calculation module obtains a real plane image i The specific process of grayscale gradient of each grid area is: gg i，j =(D g i，j -D g i，j+1 ) / Distance In the formula, Δ g i，j , Δ g i，j+1 For the i The first j indivual, j +1 The true grayscale mean of the grid area, gg i，j For the i The first j The network domain and j +1 grayscale gradient value between grid regions, Distance For the j The network domain and j +1 distance between grid regions; With the current grid ( i , j ) is the center point, calculated at Distance The grayscale gradient values ​​of all grids within this distance, and the statistical mean of all grayscale gradient values ​​is taken as the grid area j Grayscale gradient.

7. The dust control system for a storage yard according to any one of claims 1 to 6, characterized in that: A plurality of the spraying devices and camera devices are arranged in a grid pattern on the top of the storage yard.

8. The storage yard dust control system according to any one of claims 1 to 6, characterized in that: The camera device includes at least one of a gun camera, a ball camera, a gun-ball camera, and a panoramic camera.

Images