A method for detecting abnormal states in flotation machines based on markers and visual algorithms

Abstract

This invention discloses a method for detecting abnormal states of flotation machines based on markers and visual algorithms. By installing mechanical markers at the edge of the flotation machine and checking for leakage at the bends of these markers, the state of the flotation machine is determined, improving the objectivity and consistency of manual assessments. Furthermore, by using machine vision, combining the image features of the markers with the positional information of the detected features, the above process is automated, achieving higher detection accuracy than purely visual algorithms. Even when the flotation machine's foam frame is contaminated, it still provides stable and effective detectable visual features, offering higher accuracy and consistency than purely visual algorithms, and better timeliness and effectiveness than manual inspection. This invention, by monitoring the flotation machine's operating status in real time, is unaffected by pulp foam and allows for timely control and adjustment of the flotation machine, thereby stabilizing mineral processing production and optimizing mineral processing indicators.

Description

Technical Field

[0001] This invention belongs to the field of flotation machine technology, specifically relating to a processing method for detecting the working status of a self-overflowing flotation machine. Background Technology

[0002] In the mineral flotation process, whether an overflow or drop-off occurs during the operation of a self-flowing flotation machine is a crucial factor affecting the control and optimization of the flotation machine. Timely detection of these abnormalities, coupled with automatic control adjustments, is of great significance for stabilizing mineral processing production and optimizing beneficiation indicators. However, there is currently no effective method for detecting and determining whether a flotation machine is operating abnormally.

[0003] Existing methods for detecting abnormal operating conditions in flotation machines generally rely on human observation, which cannot achieve continuous, real-time online monitoring, nor can they ensure accurate and consistent judgment of abnormal operating conditions. While existing detection instruments, such as various level gauges, can effectively detect the liquid level in the flotation machine, abnormal operating conditions such as overflowing or falling into the flotation tank are directly related to the height of the foam surface, not the liquid level, and the liquid level and foam surface height are not directly related. Therefore, such instruments cannot achieve online monitoring of abnormal operating conditions in flotation machines. Furthermore, if machine vision technology is used solely to determine whether the flotation machine is in an abnormal operating condition based on foam image information, the adhesion, continuity, and irregularity of the slurry foam cause the image processing algorithm to encounter unclear image features during the recognition process, resulting in an ambiguous state between normal and abnormal conditions, making accurate judgment of the flotation machine's operating status impossible.

[0004] Therefore, to address the aforementioned shortcomings, a method is needed that is not affected by foam and can accurately and effectively determine the operating status of the flotation machine. Summary of the Invention

[0005] In view of this, and to address the shortcomings of the prior art, the present invention aims to provide a method for detecting abnormal states of flotation machines based on markers and visual algorithms. By installing mechanical markers at the edge of the flotation machine, the state of the flotation machine is determined by whether the markers are exposed at bends, thus improving the objectivity and consistency of manual assessments of the flotation machine's state. Based on these markers, the method can still provide stable and effective detectable visual features with higher accuracy even when the flotation machine's foam frame is contaminated.

[0006] A method for detecting abnormal states in a flotation machine based on markers and visual algorithms includes the following steps:

[0007] Step 1: Install mechanical markers at the edge of the flotation machine drum wall;

[0008] Step 2: Adjust the installation position of the marker on the flotation machine to ensure that the marker is within the field of view of the camera and has a suitable image size and installation position;

[0009] Step 3: After the sign is installed, the inner stepped bend and the outer stepped upper surface of the mechanical sign will exhibit stable visual characteristics.

[0010] Step 4: When the flotation machine is in an abnormal drop-in state, record the visual characteristics of the stepped bend on the inner side of the mechanical marker in this state, as well as the exact position of the mechanical marker within the camera's field of view at this time.

[0011] Step 5: With the flotation machine in normal operation, record the visual characteristics of the outer stepped upper surface of the mechanical marker in this state, as well as the exact position of the mechanical marker within the camera's field of view at this time;

[0012] Step 6: Inspect the flotation machine in operation and acquire images, extracting the visual features of the inner bend and the outer stepped upper surface of the mechanical markers from the images;

[0013] Step 7: If the visual features of the inner bend of the marker are successfully extracted in Step 6, and the extracted visual feature position coincides with the pre-extracted visual feature position in Step 4, then the flotation machine is determined to be in an abnormal state of falling into the tank.

[0014] If the visual features of the outer stepped upper surface of the marker are successfully extracted in step six, and the extracted visual feature position coincides with the pre-extracted visual feature position in step five, then the flotation machine is determined to be in normal condition.

[0015] Otherwise, it is in a slotting state;

[0016] Step 8: Extract a new frame of global image from the machine vision system, return to step 6, and based on the above steps, realize real-time automatic detection of abnormal states of the flotation machine.

[0017] Furthermore, step seven includes the following steps:

[0018] A: Using machine vision algorithms, with the image of the inner stepped bend of the marker in the trough state extracted in step four as a template, we perform matching in each image of the machine vision system to find the image position X1 with the maximum matching degree.

[0019] B: Extract the image position X1 with the maximum matching degree in step A. If this position coincides with the visual feature position extracted in step four, it can be determined that the bend of the mechanical marker is exposed. It is then considered that the flotation machine pulp is too low and is in an abnormal state of falling into the tank.

[0020] If the location does not coincide with the location extracted in step four, it can be determined that the bend of the mechanical marker is not exposed and the flotation machine is in normal condition.

[0021] C: Using machine vision algorithms, with the image of the outer stepped upper surface of the marker under normal conditions extracted in step five as a template, matching is performed in each image of the machine vision system to find the image position X2 with the maximum matching degree.

[0022] D: Extract the image position X2 with the maximum matching degree in step C. If this position coincides with the visual feature position extracted in step five, it can be determined that the upper surface of the step is indeed exposed as a mechanical marker, and the flotation machine slurry is considered to be normal and in a normal state.

[0023] If the location does not coincide with the location extracted in step five, it can be determined that the mechanical marker has been submerged by foam, and the flotation machine is in an abnormal state of overflow.

[0024] Furthermore, the mechanical marker has stepped bending parts on both the inner and outer sides, and the stepped bending parts are abutted against the edge of the flotation machine cylinder wall by screws.

[0025] Furthermore, the assembly method of the mechanical markers and the flotation machine cylinder wall can be adjusted by adjusting the length and number of screws.

[0026] Furthermore, under abnormal and normal conditions, the mechanical markers exhibit significant differences in image features, namely: the bends of the mechanical markers are significantly darker than other areas, and the absolute position of the bends in the image changes.

[0027] Furthermore, under abnormal and normal operating conditions, the mechanical markers exhibit significant differences in image features, namely, the stepped upper surface of the mechanical markers disappears from the camera's field of view.

[0028] Furthermore, the machine vision algorithm includes macroblock matching, edge detection, and texture feature extraction.

[0029] The beneficial effects of this invention are:

[0030] This invention uses mechanical markers installed on the edge of the flotation machine to determine the machine's status by observing whether the markers are exposed at the bends. This improves the objectivity and consistency of manual assessment of the flotation machine's status and offers higher detection reliability and accuracy compared to purely visual algorithms.

[0031] Based on this marker, it can still provide stable and effective detectable visual features even when the flotation machine's foam frame is contaminated, thereby enabling the detection of the flotation machine's operating status. By detecting the flotation machine's working status in real time, the flotation machine can be controlled and adjusted in a timely manner, thereby stabilizing mineral processing production and optimizing mineral processing indicators. Attached Figure Description

[0032] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. 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 these drawings without paying any creative work.

[0033] Figure 1 This is a schematic diagram of the assembled mechanical marker of the present invention;

[0034] Figure 2 This is a structural schematic diagram of another assembly method of the mechanical marker of the present invention;

[0035] Figure 3 This is a schematic diagram of the installation of the mechanical marker of the present invention on a flotation machine;

[0036] Figure 4 This is a flowchart illustrating the abnormal state detection and handling method for flotation machines provided by the present invention.

[0037] The attached diagram is labeled as follows: 1. Mechanical marker, 2. Stepped bending component, 3. Screw, 4. Flotation machine barrel wall. Detailed Implementation

[0038] Specific embodiments are given below to further clarify, completely, and in detail the technical solution of the present invention. These embodiments are the preferred embodiments based on the technical solution of the present invention, but the scope of protection of the present invention is not limited to the following embodiments.

[0039] Example 1:

[0040] By installing mechanical markers 1 along the edge of the flotation machine cylinder wall 4, the condition of the flotation machine can be determined by whether the bends and stepped upper surfaces of the markers are exposed. This allows for better objectivity and consistency in manual assessment of the flotation machine's condition. Based on these markers, stable and effective visual characteristics can still be provided even when the flotation machine's foam frame is contaminated.

[0041] The following describes examples of the present invention in further detail with reference to the accompanying drawings, such as... Figure 4 As shown, the method includes:

[0042] Step 1: As Figure 1As shown, a stepped mechanical marker 1 is installed and fixed to the edge of the flotation machine;

[0043] Step 2: Adjust the installation position of mechanical marker 1 on the flotation machine to ensure that mechanical marker 1 is within the field of view of the camera, has a suitable image size and installation position, and achieves the desired effect. Figure 3 As shown, this is to meet the detection and processing requirements of subsequent machine vision algorithms;

[0044] Step 3: Adjust the flotation machine to the trough state. Using the machine vision system, record the image features of the mechanical marker 1 inside the flotation machine in this state, as well as the exact position of the mechanical marker 1 within the camera's field of view.

[0045] Step 4: While keeping the flotation machine in normal operating condition, use the machine vision system to record the image features of the mechanical marker 1 on the outside of the flotation machine and the exact position of the mechanical marker 1 within the camera's field of view.

[0046] Step 5: By comparing the images of the flotation machine under abnormal and normal conditions, it can be found that the inner mechanical marker 1 shows obvious differences in image features, namely: the bend of the mechanical marker 1 is significantly darker than other areas, and the absolute position of the bend in the image changes.

[0047] Using the macroblock matching algorithm, the image of the marker bending point in the grooved state extracted in step 3 is used as a template. In each image of the machine vision system, the matching is performed to find the image position with the highest matching degree, that is, the image most similar to the marker bending point image in step 3.

[0048] In the specific implementation, a macroblock matching algorithm based on correlation coefficients is used:

[0049]

[0050]

[0051]

[0052] Where T represents the template image, I represents the image to be matched, x and y represent the coordinates of the top-left element of the current search box in the I matrix, and x' and y' represent the element coordinates of the matrix of T and the I matrix bounded by the search box;

[0053] Step 6: Extract the location of the maximum value in the Rcoeff image from Step 5. If this location coincides with the location extracted in Step 3, it can be determined that the bend of mechanical marker 1 is indeed exposed, and the flotation machine pulp is considered to be too low and in an abnormal state of falling into the tank. If this location does not coincide with the location extracted in Step 3, it can be determined that the bend of mechanical marker 1 is not exposed and the flotation machine is in a normal state.

[0054] Step 7: By comparing the images of the flotation machine under abnormal conditions and normal operation, it can be found that the outer mechanical marker 1 shows obvious differences in image features, namely: the stepped upper surface of mechanical marker 1 disappears from the camera's field of view.

[0055] Step 8: Using the macroblock matching algorithm, with the image of the outer marker in the normal state extracted in Step 4 as the template, match it in each image of the machine vision system to find the image position with the highest matching degree, that is, the image that is most similar to the outer stepped upper surface image of the outer marker in Step 4.

[0056] Step 9: Extract the location of the maximum matching degree in Step 8. If the location coincides with the location extracted in Step 4, it can be determined that the upper surface of the mechanical marker 1 is indeed exposed, and the flotation machine pulp is considered to be normal and in a normal state. If the location does not coincide with the location extracted in Step 4, it can be determined that the mechanical marker 1 has been submerged by foam, and the flotation machine is in an abnormal state of overflow.

[0057] Step 10: Extract a new frame of global image from the machine vision system, return to step 5, and continue to detect the flotation machine status.

[0058] Furthermore, the mechanical marker 1 has stepped bent parts 2 on both its inner and outer sides. These stepped bent parts 2 are abutted against the edge of the flotation machine cylinder wall 4 by screws 3 and nuts. Figure 1 , Figure 2 As shown.

[0059] Furthermore, a washer is provided between the screw 3 and the stepped bending member 2.

[0060] Furthermore, the bend of the stepped bending part 2 on the inner side of the mechanical marker 1 exhibits a stable visual feature, and whether the flotation machine is in an abnormal state of falling into the tank has a stable and reliable correlation with whether this visual feature appears in a suitable position in the image field of view.

[0061] The upper surface of the stepped bend 2 on the outer side of the mechanical marker 1 exhibits a stable visual feature, and whether the flotation machine is in an abnormal overflow state is reliably correlated with whether this visual feature appears in a suitable position in the image field of view.

[0062] Furthermore, the assembly method of the mechanical marker 1 and the flotation machine cylinder wall 4 can be adjusted by adjusting the length and number of screws 3.

[0063] Furthermore, the installation height of the trough detection mechanical marker 1 must be such that it covers the visual features when the flotation machine is operating normally; the installation height of the trough detection mechanical marker 1 must be such that it fully exposes the visual features when the flotation machine is operating normally.

[0064] Furthermore, the visual features mentioned above generally refer to features such as color, shape, and texture, all of which can play a role.

[0065] In summary, this invention, by installing mechanical markers 1 at the edge of the flotation machine and judging the machine's status by whether the markers are exposed at their bends, provides better objectivity and consistency for manual assessment of the flotation machine's status. It also provides stable and effective detectable visual features even when the flotation machine's foam frame is contaminated, offering higher accuracy and consistency than simple visual algorithms and better timeliness and effectiveness than manual inspection. Furthermore, by using machine vision algorithms, such as edge detection and macroblock matching, combined with the image features of the markers and the positional information of the detected features, the above process is automated, achieving higher detection accuracy than simple visual algorithms. This invention, by monitoring the flotation machine's operating status in real time, unaffected by pulp foam, allows for timely control and adjustment of the flotation machine, thereby stabilizing mineral processing production and optimizing mineral processing indicators.

[0066] The foregoing has shown and described the main features, basic principles, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention based on actual circumstances without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for detecting abnormal states of a flotation machine based on markers and visual algorithms, characterized in that, Includes the following steps: Step 1: Install mechanical markers at the edge of the flotation machine drum wall; Step 2: Adjust the installation position of the marker on the flotation machine to ensure that the marker is within the field of view of the camera and has a suitable image size and installation position; Step 3: After the sign is installed, the inner stepped bend and the outer stepped upper surface of the mechanical sign will exhibit stable visual characteristics. Step 4: When the flotation machine is in an abnormal drop-in state, record the visual characteristics of the stepped bend on the inner side of the mechanical marker in this state, as well as the exact position of the mechanical marker within the camera's field of view at this time. Step 5: With the flotation machine in normal operation, record the visual characteristics of the outer stepped upper surface of the mechanical marker in this state, as well as the exact position of the mechanical marker within the camera's field of view at this time; Step 6: Inspect the flotation machine in operation and acquire images, extracting the visual features of the inner bend and the outer stepped upper surface of the mechanical markers from the images; Step 7: If the visual features of the inner bend of the marker are successfully extracted in Step 6, and the extracted visual feature position coincides with the pre-extracted visual feature position in Step 4, then the flotation machine is determined to be in an abnormal state of falling into the tank. If the visual features of the outer stepped upper surface of the marker are successfully extracted in step six, and the extracted visual feature position coincides with the pre-extracted visual feature position in step five, then the flotation machine is determined to be in normal condition. Otherwise, it is in a slotting state; Step 8: Extract a new frame of global image from the machine vision system, return to step 6, and based on the above steps, realize real-time automatic detection of abnormal states of the flotation machine.

2. The flotation machine abnormality detection method based on markers and visual algorithms according to claim 1, characterized in that, In step seven Includes the following steps: A: Using machine vision algorithms, with the image of the inner stepped bend of the marker in the trough state extracted in step four as a template, we perform matching in each image of the machine vision system to find the image position X1 with the maximum matching degree. B: Extract the image position X1 with the maximum matching degree in step A. If this position coincides with the visual feature position extracted in step four, it can be determined that the bend of the mechanical marker is exposed. It is then considered that the flotation machine pulp is too low and is in an abnormal state of falling into the tank. If the location does not coincide with the location extracted in step four, it can be determined that the bend of the mechanical marker is not exposed and the flotation machine is in normal condition. C: Using machine vision algorithms, with the image of the outer stepped upper surface of the marker under normal conditions extracted in step five as a template, matching is performed in each image of the machine vision system to find the image position X2 with the maximum matching degree. D: Extract the image position X2 with the maximum matching degree in step C. If this position coincides with the visual feature position extracted in step five, it can be determined that the upper surface of the step is indeed exposed as a mechanical marker, and the flotation machine slurry is considered to be normal and in a normal state. If the location does not coincide with the location extracted in step five, it can be determined that the mechanical marker has been submerged by foam, and the flotation machine is in an abnormal state of overflow.

3. The flotation machine abnormality detection method based on markers and visual algorithms according to claim 1, characterized in that, The mechanical marker is a stepped bending member with both the inner and outer sides, and the stepped bending member is abutted against the edge of the flotation machine cylinder wall by screws.

4. The flotation machine abnormality detection method based on markers and visual algorithms according to claim 1, characterized in that, The assembly method of the mechanical markers and the flotation machine cylinder wall can be adjusted by adjusting the length and number of screws.

5. The method for detecting abnormal states of a flotation machine based on markers and visual algorithms according to claim 1, characterized in that, Under abnormal and normal conditions, the mechanical markers exhibit significant differences in image features, namely: the bends of the mechanical markers are significantly darker than other areas, and the absolute position of the bends in the image changes.

6. The flotation machine abnormality detection method based on markers and visual algorithms according to claim 1, characterized in that, Under abnormal and normal operating conditions, the mechanical markers exhibit significant differences in image features, namely: the stepped upper surface of the mechanical markers disappears from the camera's field of view.

7. The method for detecting abnormal states of a flotation machine based on markers and visual algorithms according to claim 2, characterized in that, The machine vision algorithm includes macroblock matching, edge detection, and texture feature extraction.

Images