Machine Vision-Based Method for Detecting Blockage in Coal Preparation Plant Screening Tanks

Abstract

This invention discloses a machine vision-based method for detecting blockages in coal preparation plant screening tanks, belonging to the field of industrial automation detection technology. The method includes: acquiring continuous video frames; the user defining an irregular polygonal monitoring area and generating a mask; dividing the video frames into equally spaced blocks; performing normalized correlation template matching on each block within the monitoring area using adjacent frames, recording the positional deviation; marking blocks with deviations close to 0 in multiple consecutive frames as continuously stationary blocks; and determining blockage and triggering an alarm when the number of stationary blocks is greater than a threshold. This invention improves robustness to changes in illumination by utilizing normalized correlation matching, effectively filters out instantaneous misjudgments through continuous frame stationary detection and an accumulator mechanism, and solves the problem of water flow interference by combining edge detection preprocessing, thus achieving automatic and accurate detection of blockages in coal preparation plant screening tanks and ensuring production safety.

Description

Technical Field

[0001] This invention relates to the field of industrial automation testing technology, and in particular to a method for detecting blockages in screening tanks of coal preparation plants based on machine vision. Background Technology

[0002] In the coal washing and processing plant, the screening trough uses the vibration of the screen plates or the movement of water to slowly move coal blocks or gangue, separating coal blocks of different sizes into layers based on the size of the holes in the screen plates. During this process, large pieces of coal or foreign objects may appear in the trough, causing partial or complete blockage of the channel. In such cases, timely manual intervention is required to clear the blocked area to ensure production safety.

[0003] In traditional coal washing plants, workers rely on visual inspection to determine if the guide channels are blocked. This method is not only inefficient but also prone to missed diagnoses. Furthermore, coal preparation plants typically have high dust levels and noise levels, posing a serious threat to workers' health if they work in such an environment for extended periods.

[0004] Existing technologies also employ image processing techniques for detection, such as the inter-frame difference method. However, the inter-frame difference method has significant drawbacks in practical applications: it is highly sensitive to environmental factors such as changes in illumination and shadows, easily leading to false positives or false negatives; furthermore, threshold selection is difficult, and its robustness to changes in illumination and noise is poor. Therefore, there is an urgent need for a blockage detection method that can adapt to the complex environment of coal washing plants, has high detection accuracy, and strong anti-interference capabilities. Summary of the Invention

[0005] The purpose of this invention is to provide a machine vision-based method for detecting blockage in screening tanks in coal preparation plants. This method utilizes machine vision technology to overcome human error, has strong environmental adaptability, and can achieve automatic and accurate detection of the blockage status of screening tanks.

[0006] To achieve the above objectives, the present invention provides a machine vision-based method for detecting blockages in screening tanks of coal preparation plants, comprising the following steps: S1. Acquire continuous video frames, capturing a single-channel grayscale video frame image, defined as... ,in Take a positive integer to represent the frame number. and Indicates the width and height of the video image; S2. Set the monitoring area. Based on the current monitoring scene, the user delineates an irregular polygon to form a closed area, and generates a binary image as a mask image based on the corner points of the polygon selected by the user. This image is denoted as S2. ; S3. Divide the video frames into equally spaced blocks, uniformly dividing the video frames into... From a cube of a certain size, we obtain sub-blocks, where... This represents the coordinates of the top-left corner of the sub-block. Indicates the width and height of the sub-block. Indicates the index of the sub-block in the row and column directions, and takes a positive integer value; S4. Perform template matching on adjacent video frames for each block within the monitoring area, using the first frame as the template. A sub-block in the frame is a template , with the first The area around the sub-block in the frame The area of ​​each sub-block is the matching area. The template is calculated using a normalized correlation matching algorithm. Match degree with the current matching position ,turn up The largest position is taken as the best matching position, and the best matching position and the sub-block are recorded at the 1st position. The deviation between the original positions in the frame; S5. Extract stationary blocks. Mark blocks with matching deviations close to 0 as stationary blocks. When sub-blocks are in continuous... When the deviations within a frame are all close to 0, it is identified as a persistent static block; S6. System alarm determination: When blocks marked as stationary are connected to each other in position to form a whole, and the total number of connected stationary blocks exceeds a set threshold. If this occurs, it is determined to be a blockage and an alarm will be triggered.

[0007] Preferably, in step S3, The value of is related to the image resolution and the minimum possible clogging area. In practical applications... The value is 16.

[0008] Preferably, in step S5, The value is ,in This refers to the video frame rate.

[0009] Preferably, in step S4, the normalized correlation matching algorithm applies the template... Match degree with the current matching position Define the following formula: ; in, The coordinates of the currently matched position. These are relative coordinates within the block, and their range depends on the size of the block. , and These represent the pixel values ​​of the template and the matching region, respectively.

[0010] Preferably, in step S5, the confirmation of the stationary block adopts an accumulator statistical method: an accumulator is set for each sub-block position, and when the matching deviation at that position is continuous... When the frame is close to 0, the accumulator accumulates to... If the deviation is not close to 0, the sub-block is confirmed to be a continuous static block; if the deviation of a frame is not close to 0, the accumulator is cleared.

[0011] Preferably, in step S4, before performing template matching, the step of performing edge detection preprocessing on the video frame is further included: performing edge detection on the image within the monitoring area, and if there is no edge information in the area, then no matching is performed or it is determined to be a non-blocked area, so as to filter out the interference caused by the water flow.

[0012] Preferably, in step S4, when performing template matching, the template is searched in a local search mode on the image to be matched, and the search range is a set area that expands outward from the current template position.

[0013] Preferably, in step S2, the user manually selects the ROI region through the interactive interface to delineate the irregular polygon.

[0014] Preferably, the method further includes a running speed optimization step: when selecting a template, the step size of the template movement is equal to the size of the template, that is, sub-blocks are selected sequentially for matching in a grid manner.

[0015] Preferably, in step S6, after the system determines that there is a blockage, it displays the location of the blockage area on the interface and sends an alarm message to the monitoring center via an IO signal.

[0016] Therefore, the machine vision-based method for detecting blockages in coal preparation plant screening tanks using the above-described structure has the following beneficial effects: (1) The present invention adopts a template matching algorithm based on normalized correlation. Compared with the traditional inter-frame difference method, it has stronger robustness to environmental interference such as illumination changes and shadows, and can effectively reduce the false detection rate and false negative rate, thereby improving the detection accuracy.

[0017] (2) This invention sets up multiple consecutive frames ( The frame-based static judgment and accumulator mechanism can distinguish between accidental static coal blocks and actual blockages, filtering out misjudgments caused by different coal blocks accidentally stopping or moving to the same position.

[0018] (3) The present invention uses block processing and connectivity analysis, which can accurately identify blockage areas with a certain area and avoid false alarms for isolated small stationary objects.

[0019] (4) This invention supports user-defined irregular polygonal monitoring areas (ROI), enabling targeted monitoring of key areas and eliminating background interference.

[0020] (5) By introducing edge detection preprocessing, the present invention effectively solves the problem of false detection that may be caused by similar texture features on smooth water flow surfaces.

[0021] (6) The method provided by the present invention realizes automated detection, replaces manual visual observation, and protects the health of workers and production safety.

[0022] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0023] Figure 1 These are schematic diagrams of the operation status of the screening tank under different scenarios of the present invention, wherein (a) is an actual schematic diagram of the normal operation status of the screening tank, (b) is an actual outline schematic diagram of the normal operation status of the screening tank, (c) is an actual schematic diagram of the operation status of the screening tank when it is blocked, and (d) is an actual outline schematic diagram of the operation status of the screening tank when it is blocked. Figure 2 This is a schematic diagram of the user-defined monitoring area and the generated binary mask image in the present invention, wherein (a) is a schematic diagram of the irregular polygonal monitoring area defined by the user in the first monitoring scenario, (b) is a schematic diagram of the generated binary mask image in the first monitoring scenario, (c) is a schematic diagram of the irregular polygonal monitoring area defined by the user in the second monitoring scenario, and (d) is a schematic diagram of the generated binary mask image in the second monitoring scenario. Figure 3 This is a schematic diagram of dividing video frames into equally spaced blocks in this invention; Figure 4 This is a schematic diagram of the template and the search area during template matching in this invention; Figure 5 This is a schematic diagram illustrating the alarm marking of potentially congested areas in this invention; Figure 6 This is a flowchart of the method provided by the present invention. Detailed Implementation

[0024] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0025] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0026] Example 1 like Figure 6 As shown in the figure, this embodiment of the invention provides a machine vision-based method for detecting blockages in coal preparation plant screening tanks. The specific steps are as follows: Step S1: Acquire consecutive video frames.

[0027] After the system starts, it acquires real-time video streams of the screening tank area via industrial cameras or monitoring equipment. For example... Figure 1 The diagram shows the operating status of the screening tank under different scenarios. Among them, Figure 1 Images (a) and (b) show the normal operating state of the screening tank, where coal blocks move under the influence of water flow; while... Figure 1 In (c) and (d) of the diagram, large foreign objects or large pieces of coal can be seen inside the screening tank. These objects remain inside the tank, causing localized blockages. This method aims to automatically identify the blockage scenarios shown in the diagram.

[0028] To reduce data processing volume, captured video frame images need to be processed into single-channel grayscale images. Define the... The frame grayscale image is ,in Take a positive integer to represent the frame number. and Indicates the width and height of the video image. These are pixel coordinates.

[0029] Step S2: Set the monitoring area.

[0030] Because the effective area of ​​interest varies in different monitoring scenarios (e.g., different filter positions, different shooting angles), the system provides an interactive interface that allows users to manually select the Region of Interest (ROI). Users can define closed regions on the video screen using irregular polygons. Based on the user-selected polygon corner points, the system generates a binary mask image of the same size as the video frame. .like Figure 2 As shown, Figure 2 (a) in the diagram is a schematic diagram of the irregular polygonal monitoring area defined by the user in the first monitoring scenario. This area covers the first filter channel that needs to be monitored. Figure 2 (b) is a schematic diagram of the binary mask image generated in the first monitoring scenario. Figure 2 (c) is a schematic diagram of an irregular polygonal monitoring area defined by the user in the second monitoring scenario, showing the situation of defining a filter channel at a different location or angle. Figure 2 (d) in the diagram is a schematic of the binary mask image generated in the second monitoring scenario. In the mask image generated above, the pixel value inside the polygon is 1 (white), and the value outside is 0 (black). All subsequent image processing operations are performed within the area where the mask value is 1, thereby eliminating interference from irrelevant backgrounds (such as walls, conveyor belts, or other static backgrounds or irrelevant equipment).

[0031] Step S3: Divide the video frames into equally spaced blocks.

[0032] To achieve refined analysis of image content, video frames are uniformly divided into squares of the same size within the monitored area. Specifically, the video frames are divided into... A cube of a certain size. Parameters The value of is related to the image resolution and the minimum blockage area to be detected. If Taking values ​​that are too small will lead to a sharp increase in computation and make the calculations more susceptible to noise; if If the value is too large, small-area blockages may be missed. Based on experience in actual coal washing scenarios, extremely small blockages are unlikely to occur; therefore, this embodiment preferably uses [a specific value]. .

[0033] like Figure 3 As shown, the sub-blocks are defined as follows: ,in Represents the row and column indices of the sub-block, and is a positive integer. Let the width and height of the video frame be... Then the row direction number The range of values ​​is Column direction number The range of values ​​is . sub-block The position in the corresponding image is determined by its top-left corner coordinates. Determined can be represented as .

[0034] Step S4: Perform template matching of adjacent video frames for each block within the monitoring area.

[0035] This step is the core of detecting stationary blockages. For any sub-block within the monitored area, let its position be the [number]th [block]. The position in the frame is The sub-block is in the first... Image content in the frame as a template Then, in the... In the frame, by position Centered on, select the surrounding area The region, the size of a sub-block, is used as the matching search range. .

[0036] like Figure 4 As shown, the red square represents the template currently to be matched. The green area represents the search scope. The template is calculated using a normalized correlation matching algorithm. Search scope The degree of matching within.

[0037] The formula for the normalized correlation matching algorithm is as follows: ; in, The coordinates of the current matching position. These are relative coordinates within the sub-block, and their values ​​range from the block size. Decide, That is, the average pixel value within the block needs to be subtracted during the calculation.

[0038] The search range is traversed using the above formula. Find the matching degree The position with the largest value is denoted as the optimal matching position. Then, the optimal matching position and the template are calculated. In the Displacement deviation between the original positions of the frames.

[0039] Step S5: Extract the blocks with matching deviations close to 0.

[0040] If the matching position deviation of a sub-block between two frames is close to 0, it means that the content of the sub-block has hardly moved and is in a "static" state. However, coal blocks in the screening tank may occasionally pause briefly due to the water flow, or there may be situations where coal blocks with similar appearances move to the same position, leading to momentary misjudgment.

[0041] Therefore, this embodiment introduces a time-dimensional continuity determination. Only when the sub-block is continuous... Within a frame, a block is marked as a "persistent still block" only when its matching deviation is close to 0. Parameters The settings determine the detection sensitivity and false positive rate. In practical applications, the preferred settings are... (Video frame rate). For example, if the video frame rate is 25 frames per second, then That is, a block needs to remain still for more than 5 seconds to be considered a stationary block.

[0042] Step S6: Setting system alarm conditions.

[0043] Real-world blockages typically manifest as accumulations of a certain area, rather than isolated pixels. Therefore, the algorithm further performs spatial connectivity analysis on stationary blocks. When multiple sub-blocks marked as stationary are spatially adjacent and connected to form a whole, and the total number of sub-blocks contained in this connected whole exceeds a preset threshold... At that time, the system determined that a blockage had occurred. Threshold The alarm setting can be based on the smallest actual blockage area. For example... Figure 5 As shown, the system will select and mark the detected blockage area on the video screen and issue an alarm.

[0044] Example 2 This embodiment further optimizes the implementation details of the algorithm based on Embodiment 1 to solve the problems of large computational load and false detection.

[0045] (1) Optimization of accumulator mechanism for static determination: To achieve "continuous" more precisely "Frame stillness determination" – this embodiment uses an accumulator array for statistical analysis. The specific method is as follows: Create an accumulator array corresponding to the number of sub-blocks in the ROI region. For each sub-block position, perform the following logical checks: If the matching deviation calculated in the current frame is close to 0, then the accumulator value corresponding to that position is incremented by 1; If the matching deviation of the current frame is not close to 0, the accumulator value corresponding to that position is cleared to zero; When the accumulator value reaches the set threshold (For example If the condition is met, then the sub-block is confirmed to be a continuously stationary block.

[0046] This method can effectively avoid momentary false detections caused by similar coal blocks accidentally flowing through the same location, greatly improving the accuracy of detection.

[0047] (2) Optimization of running speed: When performing template matching, selecting templates pixel-by-pixel and performing a global search would result in an extremely high computational load, potentially causing system lag or crashes. This embodiment employs the following optimization measures: Template selection step size optimization: When selecting template sub-blocks, instead of using a single pixel as the step size, the step size is the size of the sub-block (e.g., 16 pixels), meaning sub-blocks are selected sequentially in a grid pattern for matching. This ensures coverage of the entire image while significantly reducing the number of template matches.

[0048] Search range optimization: When searching within the image to be matched, instead of performing a full-image search, the search is centered on the current sub-block position and only searches the surrounding local area (e.g., expanding outwards to include...). (or a larger sub-block region). Since the blockage is mainly stationary and its movement range is extremely small, local search is sufficient to capture its positional changes and greatly reduces computation time.

[0049] Example 3 This embodiment has been adapted to specific interference scenarios.

[0050] Improvement of water flow false detection: In a coal preparation plant scenario, if the monitored area contains a gently flowing water surface without coal blocks, it may be misjudged as a static, blocked area due to the indistinct changes in water surface texture. To address this issue, this embodiment adds an edge detection preprocessing step.

[0051] The specific steps are as follows: Before template matching, edge detection is performed on the image using the filter2D function in OpenCV (e.g., using the Laplacian operator or Sobel kernel).

[0052] Calculate the edge information intensity within the current region to be matched.

[0053] If the region is found to lack edge information (such as a smooth water surface), it will not be judged as a blockage, or the subsequent static determination logic will not be executed, even if its matching deviation is close to 0.

[0054] Only in areas with abundant edge features (such as coal block edges and textures) are static features identified as potential blockage areas.

[0055] This step effectively filters out interference from the smooth water flow background, further improving the robustness of the system.

[0056] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as ROM, RAM, magnetic disk, or optical disk.

[0057] 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 preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A machine vision-based method for detecting blockages in screening tanks at coal preparation plants, characterized in that, Includes the following steps: S1. Acquire continuous video frames, capturing a single-channel grayscale video frame image, defined as... ,in Take a positive integer to represent the frame number. and Indicates the width and height of the video image; S2. Set the monitoring area. Based on the current monitoring scene, the user delineates an irregular polygon to form a closed area, and generates a binary image as a mask image based on the corner points of the polygon selected by the user. This image is denoted as S2. ; S3. Divide the video frames into equally spaced blocks, uniformly dividing the video frames into... A cube of a certain size is used to obtain sub-blocks, among which... This represents the coordinates of the top-left corner of the sub-block. This represents the width and height of the sub-block. Indicates the index of the sub-block in the row and column directions, and takes a positive integer value; S4. Perform template matching on adjacent video frames for each block within the monitoring area, using the first frame as the template. A sub-block in the frame is a template , with the first The area around the sub-block in the frame The area of ​​each sub-block is the matching area. The template is calculated using a normalized correlation matching algorithm. Match degree with the current matching position ,turn up The largest position is taken as the best matching position, and the best matching position and the sub-block are recorded at the 1st position. The deviation between the original positions in the frame; S5. Extract stationary blocks. Mark blocks with matching deviations close to 0 as stationary blocks. When sub-blocks are in continuous... When the deviations within a frame are all close to 0, it is identified as a persistent static block; S6. System alarm determination: When blocks marked as stationary are connected to each other in position to form a whole, and the total number of connected stationary blocks exceeds a set threshold. If this occurs, it is determined to be a blockage and an alarm will be triggered.

2. The method for detecting blockage in a coal preparation plant screening tank based on machine vision according to claim 1, characterized in that: In step S3, The value of is related to the image resolution and the minimum possible blockage area.

3. The method for detecting blockage in a coal preparation plant screening tank based on machine vision according to claim 1, characterized in that: In step S4, the normalized correlation matching algorithm applies the template. Match degree with the current matching position The definition is as follows: ； in, The coordinates of the currently matched position. These are relative coordinates within the block, and their range depends on the size of the block. .

4. The method for detecting blockage in a coal preparation plant screening tank based on machine vision according to claim 1, characterized in that: In step S4, before template matching, the video frame is preprocessed by edge detection: edge detection is performed on the image within the monitoring area. If there is no edge information in the area, no matching is performed or the area is determined to be a non-blocked area to filter out the interference caused by water flow.

5. The method for detecting blockage in a coal preparation plant screening tank based on machine vision according to claim 1, characterized in that: In step S4, when performing template matching, the template is selected sequentially according to the sub-block size, and the search method is a local search centered on the current sub-block position.

6. The method for detecting blockage in a coal preparation plant screening tank based on machine vision according to claim 1, characterized in that: In step S5, the confirmation of stationary blocks uses an accumulator statistical method: an accumulator is set for each sub-block position, and when the matching deviation at that position is continuous... When the frame is close to 0, the accumulator accumulates to... If the deviation is not close to 0, the sub-block is confirmed to be a continuous static block; if the deviation of a frame is not close to 0, the accumulator is cleared.

7. The method for detecting blockage in a coal preparation plant screening tank based on machine vision according to claim 1, characterized in that: In step S5, The value is ,in This refers to the video frame rate.

8. The method for detecting blockage in a coal preparation plant screening tank based on machine vision according to claim 1, characterized in that: In step S6, after the system determines that there is a blockage, it displays the location of the blockage area on the interface and sends an alarm message to the monitoring center via an IO signal.

9. The method for detecting blockage in a coal preparation plant screening tank based on machine vision according to claim 1, characterized in that: In step S2, the user manually selects the ROI region through the interactive interface to delineate the irregular polygon.

10. The method for detecting blockage in a coal preparation plant screening tank based on machine vision according to claim 1, characterized in that: It also includes step S7: when the detected stationary blocks are connected to each other and the area characteristics of the connected domain meet the blockage characteristics, output a blockage alarm and location coordinates.

Images