Drum experiment ice rock sorting degree automatic measurement method, system and equipment based on visual identification and storage medium
By using top-down imaging and ice particle area ratio as indicators, the problem of real-time measurement of ice-rock sorting degree was solved, achieving highly stable and continuous sorting degree monitoring, which is applicable to ice-rock sorting experiments and the analysis of other mixed systems.
Patent Information
- Application Number
- CN202511106021.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-08
- Publication Date
- 2025-11-11
AI Technical Summary
Existing technologies cannot accurately measure the sorting degree of ice and rock particles in real experiments, cannot achieve real-time monitoring and data continuity, and image recognition is affected by sidewall wear and water film interference, resulting in large measurement errors.
A visual recognition-based drum experiment method was adopted. Images were acquired through top-down imaging, and image preprocessing and pixel statistics of ice and rock particles were performed. The sorting degree was calculated in real time, and the area ratio of ice particles was used as the sorting degree index to establish a standardized calculation model between 0 and 1.
It achieves fully automated, continuous, and real-time monitoring of the ice-rock sorting process, avoiding the lag and errors of traditional methods, improving the stability of image recognition and the reliability of data, and is suitable for the intelligent upgrading of experimental platforms.
Smart Images
Figure CN120927546A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of geological disaster simulation and testing technology, and in particular relates to an automatic measurement method, system, equipment and storage medium for ice and rock sorting degree in a rotating drum experiment based on visual recognition. Background Technology
[0002] In the study of geological hazards, especially cold-region hazards such as ice-rock debris flows and glacial debris flows, sorting behavior is one of the key physical processes and mechanisms influencing flow patterns and depositional morphologies. Ice-rock mixture sorting refers to the migration and separation of ice and rock particles during flow due to differences in particle size, density, and friction, manifested as a trend of ice particles gradually rising and rock particles settling. This process not only affects the rheological characteristics, drag characteristics, and velocity characteristics of debris flows but also directly determines the sedimentary structure of the hazard deposit, its impact range, and its interaction mechanism with the affected structures. Therefore, studying the particle sorting process and quantifying the evolution characteristics of sorting degree are crucial technical steps in revealing the dynamic mechanisms of such hazards and predicting their evolutionary trends. It is also a key step in quantitatively analyzing the potential hazard losses caused by ice-rock debris flows to buildings and other affected structures, holding significant importance in both basic research and engineering applications.
[0003] Currently, quantitative research on the sorting behavior of ice and rock particles mainly employs the following three types of technical methods: monitoring methods based on numerical simulation, observation methods based on flume experiments, and long-process simulation methods based on drum experiments. However, these methods cannot accurately measure and extract particle information in real experiments, thus making it impossible to accurately measure the sorting degree in real time. Summary of the Invention
[0004] The purpose of this invention is to provide a method, system, device, and storage medium for automatic measurement of ice and rock sorting degree in a rotating drum experiment based on visual recognition, so as to solve the problem that particle information cannot be accurately measured and extracted in real experiments, thus making it impossible to accurately measure the sorting degree in real time.
[0005] The embodiments of this application are implemented as follows: An automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition includes the following steps: Image data acquisition; Image preprocessing; Identification of ice and rock particles in preprocessed images and pixel statistics of ice and rock particles; Sorting degree calculation, with real-time output of results; Specifically, the pixel area of the pixels identified as ice particles in the image is statistically analyzed and denoted as 𝐴. ice The pixel area of each pixel identified as a rock particle in the image is statistically analyzed and denoted as 𝐴. rock The total effective area of the image's field of view is denoted as 𝐴. totalLet the sorting degree K be the proportion of ice particle area, and the calculation formula is as follows: .
[0006] Optionally, in some embodiments of this application, image data acquisition includes the following steps: Install the rotary drum experimental system; Start the drum test system and perform real-time image acquisition.
[0007] Optionally, in some embodiments of this application, the drum experimental system includes: A rotating drum device includes an annular rotating cylinder with an opening on its inner surface. A guard arm is fixed to the rotating cylinder and extends to the center point of the rotating cylinder. A support frame is located on both sides of the rotating drum device. A horizontal platform is provided on the support frame, and a motor is provided on the platform. The motor is connected to the guard arm at the center point of the rotating drum. A camera is located at the bottom of the platform and is aligned with the opening of the rotating drum.
[0008] Optionally, in some embodiments of this application, at least two guard arms are provided, and a plurality of guard arms are evenly distributed on the rotating drum along the circumference of the rotating drum, with each guard arm extending to the center point of the rotating drum and converging to form a connecting portion; and / or A flow velocity and depth sensor is installed at the bottom of the platform, and a pressure impact sensor is installed at the bottom of the rotating drum. The rotating drum experimental system also includes a control system, which is electrically connected to the flow velocity and depth sensor, the pressure impact sensor, and the motor, respectively; and / or At least one side of the rotating cylinder is made of transparent material.
[0009] Optionally, in some embodiments of this application, before the experiment begins, the camera establishes a connection with the control system and initiates real-time image acquisition through the control program of the control system.
[0010] Optionally, in some embodiments of this application, image preprocessing includes performing boundary cropping, grayscale conversion, and binarization on each frame of the image; and / or Pixel statistics include counting the number of pixels for ice particles and rock particles.
[0011] Optionally, in some embodiments of this application, after each frame of image preprocessing, the sorting degree K of the current frame is output in real time. 𝑡 It also automatically generates a time-sorting curve.
[0012] Accordingly, embodiments of this application also provide an automatic measurement system for ice and rock sorting degree in a drum test based on visual recognition, including: Image data acquisition module, used for image data acquisition; Image preprocessing module, used for image preprocessing; The recognition and pixel statistics module is used for the recognition of ice particles and rock particles in the preprocessed image, as well as the pixel statistics of ice particles and rock particles. The real-time output results module is used for sorting degree calculation and outputs the results in real time. Specifically, the pixel area of the pixels identified as ice particles in the image is statistically analyzed and denoted as 𝐴. ice The pixel area of each pixel identified as a rock particle in the image is statistically analyzed and denoted as 𝐴. rock The total effective area of the image's field of view is denoted as 𝐴. total Let the sorting degree K be the proportion of ice particle area, and the calculation formula is as follows: .
[0013] Accordingly, embodiments of this application also provide a computer device, including a storage device and a processor, wherein the storage device stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.
[0014] Accordingly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.
[0015] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: This application proposes a novel method for real-time measurement and calculation of sorting degree based on drum experiments, effectively avoiding the bottleneck of traditional methods in accurately measuring and extracting spatial information of ice and rock particles in real time. By changing the imaging angle and imaging area, a top-down imaging method is innovatively adopted, avoiding the scratch interference, water film reflection, and boundary effect problems commonly found in traditional sidewall imaging. This significantly improves image quality and observation representativeness, enabling the system to maintain high stability and repeatability in long-term, multi-round experiments, and allowing for real-time quantification of image observation and automatic measurement. This application establishes a complete processing method, including real-time image data acquisition, preprocessing, ice / rock classification and pixel statistics, sorting degree calculation and output, realizing fully automatic, continuous, and real-time monitoring of the ice / rock sorting process and sorting degree indicators. This application avoids the lag of offline processing required by existing technologies, possesses a high degree of automation and closed-loop experimental data, and can repeatedly and stably obtain continuous sorting degree data, making it particularly suitable for the intelligent upgrade needs of experimental platforms.
[0016] This application is the first to use "ice particle area ratio" as a sorting index and establishes a standardized calculation model that continuously varies between 0 and 1, possessing clear physical meaning and stable mathematical properties. Compared to traditional methods that rely on indirect indicators such as changes in the center of mass position, this effectively solves the bottleneck of inaccurate particle information measurement in experiments. This index not only has clearer physical meaning and statistical stability but can also accurately calculate the sorting index in real time, significantly improving experimental adaptability. Attached Figure Description
[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings of the embodiments will be briefly described below. Obviously, the drawings described below only relate to some embodiments of this application and are not intended to limit this application, wherein: Figure 1 This is a structural diagram of the drum experimental system of the present invention; Figure 2 This is a cross-sectional view of the rotating drum where the material is located according to the present invention; Figure 3 The image preprocessing effect diagram provided by this invention; Figure 4 This is a set of test results from a sorting experiment, which is an application example of the present invention.
[0018] Explanation of reference numerals in the attached figures: 1-Control system, 2-Guard arm, 3-Rotating drum, 4-Motor, 5-Support, 6-Camera, 7-Pressure impact sensor, 8-Flow velocity and depth sensor, 9-Rock particles, 10-Ice particles. Detailed Implementation
[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.
[0020] Although existing research attempts to obtain image information of key moments in the ice-rock particle sorting process by installing a camera system on the side wall of the rotating drum experimental device, this method is consistent with the principle of numerical simulation (numerical simulation calculates the center of gravity position by extracting the three-dimensional information of all particles, while the experimental side wall photography method uses the side to represent three-dimensional spatial information). The ice and rock areas are distinguished by taking pictures, and then the ice and rock particle areas need to be manually selected with the photos. Subsequently, the center of gravity position of the area needs to be calculated, and finally the sorting degree is calculated based on the center of gravity position.
[0021] However, this method still has the following key drawbacks in practical applications: 1) Image quality is limited by physical wear on the sidewalls, making long-term stable use impossible: During the drum experiment, ice and rock particles move continuously within the drum and frequently impact or rub against the sidewalls. As the number of experiments increases, the transparent glass or acrylic sidewalls are prone to surface wear phenomena such as scratches and fogging, severely reducing image acquisition clarity and affecting the accuracy of image recognition.
[0022] 2) Melting ice particles produce liquid water that interferes with imaging quality: During the experiment, ice particles gradually melt due to friction and localized temperature rise, causing liquid water to adhere to the sidewalls of the drum, forming water films, fog spots, or reflective areas. In some cases, it mixes with fine sand and clay to form mud that adheres to the sidewalls, making measurement impossible. These interferences significantly reduce the quality of particle boundary recognition, thereby affecting the accuracy of area extraction and sorting degree calculation.
[0023] 3) Real-time continuous monitoring of the experimental process is not possible, and data updates are delayed: Existing image acquisition and processing are mostly offline operations performed after the experiment, which cannot reflect the dynamic evolution of the ice-rock sorting process in real time, thus limiting the research on the continuity and timeliness of sorting patterns.
[0024] 4) The sidewall effect exists, and the observation results lack representativeness: Due to the frictional differences and packing structure deviations between the particles and the sidewalls in the drum, the areas near the sidewalls often exhibit different motion behaviors and sorting trends than the internal particles, forming the so-called "sidewall effect". Images acquired through the sidewalls only reflect the particle state in the areas near the walls, making it difficult to truly reflect the representative sorting situation of the overall ice-rock mixture, thus affecting the scientific validity and generalization value of the experimental results.
[0025] 5) Manual photo processing is labor-intensive, can only be performed at specific times, and makes it difficult to obtain continuous sorting data. Furthermore, the analysis results are highly subjective, inaccurate, and prone to significant errors.
[0026] The technical solution of this application is as follows: This application provides an automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition, including the following steps: S01, Image Data Acquisition; S02, Image Preprocessing; S03. Identification of ice particles 10 and rock particles 9 in the preprocessed image and pixel statistics of ice particles 10 and rock particles 9. S04. Sorting degree calculation, with real-time output of results; Specifically, the pixel area of each pixel identified as ice particle 10 in the image is statistically analyzed and denoted as 𝐴. ice The pixel area of each pixel identified as rock particle 9 in the image is statistically analyzed and denoted as 𝐴. rockThe total effective area of the image's field of view is denoted as 𝐴. total Let the sorting degree K be the area ratio of ice particles of 10, and the calculation formula is as follows: .
[0027] This application establishes a complete processing method, including real-time image data acquisition, preprocessing, identification of ice particles 10 and rock particles 9, pixel statistics of ice particles 10 and rock particles 9, sorting degree calculation and output, realizing fully automatic, continuous and real-time monitoring of the ice-rock sorting process. This application avoids the lag of offline processing required by existing technologies, has a high degree of automation and closed-loop experimental data, and can repeatedly and stably obtain continuous sorting degree data, making it particularly suitable for the intelligent upgrade needs of experimental platforms.
[0028] This application is the first to use the surface "ice particle area ratio" as a sorting index and establishes a standardized calculation model that continuously varies between 0 and 1, possessing clear physical meaning and stable mathematical properties. Compared to traditional methods that rely on indirect indicators such as changes in the center of mass position (traditional indicators such as changes in the center of mass position can only be accurately statistically analyzed in numerical simulations, requiring the position and weight information of each particle, which is impossible to measure in experiments; even the side window method can only estimate through area, but the side has boundary effects and cannot represent the true internal situation), this application effectively solves the bottleneck of not being able to accurately measure and extract particle information in experiments. This index not only has a clearer physical meaning and statistical stability, but also significantly improves the accuracy of index calculation and experimental adaptability.
[0029] The image acquisition and drum experimental system constructed in this application can realize real-time image input, particle recognition, and sorting degree calculation during the experiment. It supports continuous sampling at the millisecond level (depending on the camera's 6 frames per second) and automatically generates a "time-sorting degree" change curve. This automated processing flow greatly improves experimental efficiency, avoids errors caused by repetitive manual processing, and eliminates the drawbacks of traditional offline processing, such as long processing time and data lag. It is particularly suitable for conducting dynamic process observation and parameter sensitivity analysis.
[0030] This application is not only applicable to the sorting experiments of ice-rock particles, but can also be extended to the analysis of the sorting behavior of mixture systems such as snow-rock, sand-gravel, and soil-ice. The application features a modular design in structure and adjustable parameters in software, possessing good versatility and portability, making it suitable for widespread use in teaching, research, and engineering simulation platforms.
[0031] In S01: In some embodiments, image data acquisition includes the following steps: S011. Install the rotary drum test system; S012. Start the drum test system and perform real-time image acquisition.
[0032] In S011: Please see Figure 1 and Figure 2 In some embodiments, the drum experimental system includes: The rotating drum device includes an annular rotating cylinder 3, with an opening on the inner surface of the rotating cylinder 3, and a guard arm 2 fixed on the rotating cylinder 3, the guard arm 2 extending to the center point of the rotating cylinder 3; The bracket 5 is located on both sides of the rotating drum device. A horizontal platform is provided on the bracket 5, and a motor 4 is provided on the platform. The motor 4 is connected to the guard arm 2 at the center point of the rotating drum 3. Camera 6 is located at the lower part of the platform and is aligned with the opening of the rotating drum 3.
[0033] Furthermore, at least two guard arms 2 are provided, and multiple guard arms 2 are evenly distributed on the rotating drum 3 along its circumference. Each guard arm 2 extends to the center point of the rotating drum 3 and converges to form a connecting part. It can be understood that the motor 4 drives the connecting part to rotate, thereby driving the guard arm 2 to rotate together with the rotating drum 3.
[0034] For example, the rotating cylinder 3 can be an annular rotating cylinder 3 formed by bending a U-shaped groove, with the opening of the U-shaped groove located on the inner side of the rotating cylinder 3.
[0035] As another example, the rotating cylinder 3 is a hollow annular cylinder, with an opening on the inner side of the hollow annular cylinder to form the rotating cylinder 3.
[0036] For example, there are three guard arms 2, which are evenly distributed on the rotating drum 3 along the circumference of the rotating drum 3. This provides stable support for the rotating drum 3 while facilitating the rotation of the guard arms 2 and the rotating drum 3 by the motor 4.
[0037] Furthermore, a flow velocity and depth sensor 8 is provided at the bottom of the platform, and a pressure impact sensor 7 is provided at the bottom of the rotating drum 3. The rotating drum experimental system also includes a control system 1, which is electrically connected to the flow velocity and depth sensor 8, the pressure impact sensor 7, and the motor 4, respectively.
[0038] Furthermore, at least one side of the rotating drum 3 is made of a transparent material. It is understood that making the side of the rotating drum 3 transparent facilitates observation of the rotation of the material inside the drum 3.
[0039] For example, the outer diameter of the rotating drum 3 is 200 cm, the inner width of the opening of the rotating drum 3 is 30 cm, the material is smooth and scratch-resistant high-strength acrylic, and the rotation speed of the drum device is 5 rpm to 20 rpm to simulate the movement and sorting process of the ice-rock debris mixture under different speed conditions.
[0040] It is understandable that camera 6 is fixed to the lower part of the platform with screws, and the field of view is changed by adjusting the focal length to ensure that it covers the entire top projection area of the rotating drum device.
[0041] It is understandable that after the material is loaded into the rotating drum 3, the material will accumulate at the bottom of the rotating drum 3 under the action of gravity. After the motor 4 is started, the motor 4 drives the rotating drum 3 and the material to rotate. However, due to the influence of gravity, the material will continue to slide down. During the rotation of the rotating drum 3, the material slides down like a landslide, thus forming a debris flow.
[0042] Furthermore, camera 6 is equipped with at least one.
[0043] It is understandable that multiple camera arrays can be set up to cover a larger area or provide imaging from different angles, thereby achieving higher spatial resolution.
[0044] Furthermore, ring-shaped LED cold light sources are installed on both sides of the drum device. These ring-shaped LED cold light sources on both sides of the drum device can prevent light shading and reflection interference from particles.
[0045] In S012: In some embodiments, before the experiment begins, the camera 6 establishes a connection with the control system 1 and starts real-time image acquisition through the control program of the control system 1.
[0046] It is understandable that the acquisition interval is set according to the experimental requirements, and the image data is saved to the local cache or memory in real time.
[0047] It is understandable that when acquiring image data, the camera 6 is used to adjust the focus and orientation so that the material is in the center of the image.
[0048] For example, during image data acquisition, the camera 6 is used to adjust the focal length and orientation so that the material is centered in the image.
[0049] In S02: Please see Figure 3 In some embodiments, image preprocessing includes performing boundary cropping, grayscale conversion, and binarization on each frame of the image.
[0050] It is understandable that the boundary range of image data can be manually defined to extract the target area containing the material.
[0051] Furthermore, grayscale processing includes converting a color image into a grayscale image; Binarization processing includes segmenting the grayscale image into ice particles (10) and rock particles (9) based on an adaptive thresholding algorithm.
[0052] Figure 3 In the image, ① is the original photo, ② is the result after boundary cropping, ③ is the result after desaturation, and ④ is the result after binarization.
[0053] In S03: Pixel statistics include counting the number of ice particles (10 pixels) and rock particles (9 pixels).
[0054] In S04: In some embodiments, the pixel area of the region in the image identified as ice particles 10 is statistically analyzed and denoted as 𝐴. ice The pixel area of each pixel identified as rock particle 9 in the image is statistically analyzed and denoted as 𝐴. rock The total effective area of the image's field of view is denoted as 𝐴. total Let the sorting degree K be the area ratio of ice particles of 10, and the calculation formula is as follows:
[0055] It can be understood that the total effective area of the image's field of view is the total number of pixels within the fixed boundary minus the occlusion and edge loss.
[0056] It is understandable that, 𝑆∈[0,1], in the early stage of sorting (ice particles 10 are located at the bottom layer and completely covered by rock particles 9, with no ice particles 10 on the surface), 𝑆≈0; in the later stage of sorting (ice particles 10 almost completely float to the surface), 𝑆≈1. This indicator changes continuously, has clear dimensions, and possesses physical interpretability and statistical stability.
[0057] Furthermore, after preprocessing each frame of image, the sorting degree K of the current frame is output in real time. 𝑡 It also automatically generates a time-sorting curve.
[0058] For example, the sorting degree K of the current frame is output in real time. 𝑡 Save as CSV or Excel format for easy statistical modeling.
[0059] It is understandable that time-sorting curves can be used for later analysis of the sorting process.
[0060] This application uses a top-view camera instead of traditional side-view imaging, effectively avoiding the area of frequent friction between ice and rock particles and the side wall of the rotating drum device. This avoids image blurring caused by scratches or fogging of the glass or acrylic side wall. At the same time, the top transparent observation window is in a non-contact area, which can continuously provide high-quality image acquisition during long-term, multi-round experiments, ensuring the stability and repeatability of image recognition and subsequent analysis.
[0061] The liquid water produced by the gradual melting of ice particles 10 during movement easily forms a water film or reflective layer on the sidewalls, severely interfering with image segmentation and particle boundary recognition. This application, by acquiring images from a top-down view, can effectively avoid areas of liquid water interference, greatly improving image recognition clarity and algorithm processing accuracy.
[0062] Due to limited space and frictional differences, particles near the drum wall often exhibit atypical distributions (i.e., sidewall effect). This application acquires top-down images of the entire particle field, covering the particle state in the middle and overall surface of the drum device. The extracted sorting data has stronger spatial representativeness and experimental reliability.
[0063] Secondly, embodiments of this application provide an automatic measurement system for ice and rock sorting degree in a rotating drum test based on visual recognition, comprising: Image data acquisition module, used for image data acquisition; Image preprocessing module, used for image preprocessing; The identification and pixel statistics module is used to identify ice particles 10 and rock particles 9 in the preprocessed image and to perform pixel statistics on ice particles 10 and rock particles 9. The real-time output results module is used for sorting degree calculation and outputs the results in real time. Specifically, the pixel area of each pixel identified as ice particle 10 in the image is statistically analyzed and denoted as 𝐴. ice The pixel area of each pixel identified as rock particle 9 in the image is statistically analyzed and denoted as 𝐴. rock The total effective area of the image's field of view is denoted as 𝐴. total Let the sorting degree K be the area ratio of ice particles of 10, and the calculation formula is as follows: .
[0064] In the image data acquisition module: In some embodiments, image data acquisition includes the following steps: S011. Install the rotary drum test system; S012. Start the drum test system and perform real-time image acquisition.
[0065] In S011: Please see Figure 1 and Figure 2 In some embodiments, the drum experimental system includes: The rotating drum device includes an annular rotating cylinder 3, with an opening on the inner surface of the rotating cylinder 3, and a guard arm 2 fixed on the rotating cylinder 3, the guard arm 2 extending to the center point of the rotating cylinder 3; The bracket 5 is located on both sides of the rotating drum device. A horizontal platform is provided on the bracket 5, and a motor 4 is provided on the platform. The motor 4 is connected to the guard arm 2 at the center point of the rotating drum 3. Camera 6 is located at the lower part of the platform and is aligned with the opening of the rotating drum 3.
[0066] Furthermore, at least two guard arms 2 are provided, and multiple guard arms 2 are evenly distributed on the rotating drum 3 along its circumference. Each guard arm 2 extends to the center point of the rotating drum 3 and converges to form a connecting part. It can be understood that the motor 4 drives the connecting part to rotate, thereby driving the guard arm 2 to rotate together with the rotating drum 3.
[0067] For example, the rotating cylinder 3 can be an annular rotating cylinder 3 formed by bending a U-shaped groove, with the opening of the U-shaped groove located on the inner side of the rotating cylinder 3.
[0068] As another example, the rotating cylinder 3 is a hollow annular cylinder, with an opening on the inner side of the hollow annular cylinder to form the rotating cylinder 3.
[0069] For example, there are three guard arms 2, which are evenly distributed on the rotating drum 3 along the circumference of the rotating drum 3. This provides stable support for the rotating drum 3 while facilitating the rotation of the guard arms 2 and the rotating drum 3 by the motor 4.
[0070] Furthermore, a flow velocity and depth sensor 8 is provided at the bottom of the platform, and a pressure impact sensor 7 is provided at the bottom of the rotating drum 3. The rotating drum experimental system also includes a control system 1, which is electrically connected to the flow velocity and depth sensor 8, the pressure impact sensor 7, and the motor 4, respectively.
[0071] Furthermore, at least one side of the rotating drum 3 is made of a transparent material. It is understood that making the side of the rotating drum 3 transparent facilitates observation of the rotation of the material inside the drum 3.
[0072] For example, the outer diameter of the rotating drum 3 is 200 cm, the inner width of the opening of the rotating drum 3 is 30 cm, the material is smooth and scratch-resistant high-strength acrylic, and the rotation speed of the drum device is 5 rpm to 20 rpm to simulate the movement and sorting process of the ice-rock debris mixture under different speed conditions.
[0073] It is understandable that camera 6 is fixed to the lower part of the platform with screws, and the field of view is changed by adjusting the focal length to ensure that it covers the entire top projection area of the rotating drum device.
[0074] It is understandable that after the material is loaded into the rotating drum 3, the material will accumulate at the bottom of the rotating drum 3 under the action of gravity. After the motor 4 is started, the motor 4 drives the rotating drum 3 and the material to rotate. However, due to the influence of gravity, the material will continue to slide down. During the rotation of the rotating drum 3, the material slides down like a landslide, thus forming a debris flow.
[0075] Furthermore, camera 6 is equipped with at least one.
[0076] It is understandable that multiple camera arrays can be set up to cover a larger area or provide imaging from different angles, thereby achieving higher spatial resolution.
[0077] Furthermore, ring-shaped LED cold light sources are installed on both sides of the drum device. These ring-shaped LED cold light sources on both sides of the drum device can prevent light shading and reflection interference from particles.
[0078] In S012: In some embodiments, before the experiment begins, the camera 6 establishes a connection with the control system 1 and starts real-time image acquisition through the control program of the control system 1.
[0079] It is understandable that the acquisition interval is set according to the experimental requirements, and the image data is saved to the local cache or memory in real time.
[0080] It is understandable that when acquiring image data, the camera 6 is used to adjust the focus and orientation so that the material is in the center of the image.
[0081] For example, during image data acquisition, the camera 6 is used to adjust the focal length and orientation so that the material is centered in the image.
[0082] In the image preprocessing module: Please see Figure 3 In some embodiments, image preprocessing includes performing boundary cropping, grayscale conversion, and binarization on each frame of the image.
[0083] It is understandable that the boundary range of image data can be manually defined to extract the target area containing the material.
[0084] Furthermore, grayscale processing includes converting a color image into a grayscale image; Binarization processing includes segmenting the grayscale image into ice particles (10) and rock particles (9) based on an adaptive thresholding algorithm.
[0085] Figure 3 In the image, ① is the original photo, ② is the result after boundary cropping, ③ is the result after desaturation, and ④ is the result after binarization.
[0086] In the recognition and pixel statistics module: Pixel statistics include counting the number of ice particles (10 pixels) and rock particles (9 pixels).
[0087] In the real-time output result module: In some embodiments, the pixel area of the region in the image identified as ice particles 10 is statistically analyzed and denoted as 𝐴. ice The pixel area of each pixel identified as rock particle 9 in the image is statistically analyzed and denoted as 𝐴. rock The total effective area of the image's field of view is denoted as 𝐴. total Let the sorting degree K be the area ratio of ice particles of 10, and the calculation formula is as follows:
[0088] It can be understood that the total effective area of the image's field of view is the total number of pixels within the fixed boundary minus the occlusion and edge loss.
[0089] It is understandable that, 𝑆∈[0,1], in the early stage of sorting (ice particles 10 are located at the bottom layer and completely covered by rock particles 9, with no ice particles 10 on the surface), 𝑆≈0; in the later stage of sorting (ice particles 10 almost completely float to the surface), 𝑆≈1. This indicator changes continuously, has clear dimensions, and possesses physical interpretability and statistical stability.
[0090] Furthermore, after preprocessing each frame of image, the sorting degree K of the current frame is output in real time. 𝑡 It also automatically generates a time-sorting curve.
[0091] For example, the sorting degree K of the current frame is output in real time. 𝑡 Save as CSV or Excel format for easy statistical modeling.
[0092] It is understandable that time-sorting curves can be used for later analysis of the sorting process.
[0093] This application uses a top-view camera instead of traditional side-view imaging, effectively avoiding the area of frequent friction between ice and rock particles and the side wall of the rotating drum device. This avoids image blurring caused by scratches or fogging of the glass or acrylic side wall. At the same time, the top transparent observation window is in a non-contact area, which can continuously provide high-quality image acquisition during long-term, multi-round experiments, ensuring the stability and repeatability of image recognition and subsequent analysis.
[0094] The liquid water produced by the gradual melting of ice particles 10 during movement easily forms a water film or reflective layer on the sidewalls, severely interfering with image segmentation and particle boundary recognition. This application, by acquiring images from a top-down view, can effectively avoid areas of liquid water interference, greatly improving image recognition clarity and algorithm processing accuracy.
[0095] Due to limited space and frictional differences, particles near the drum wall often exhibit atypical distributions (i.e., sidewall effect). This application acquires top-down images of the entire particle field, covering the particle state in the middle and overall surface of the drum device. The extracted sorting data has stronger spatial representativeness and experimental reliability.
[0096] Thirdly, this application provides a computer device including a storage device and a processor. The storage device stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the above-described automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition.
[0097] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.
[0098] The memory includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or D-interface display memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory may be an internal storage unit of the computer device, such as the hard disk or memory of the computer device. In other embodiments, the memory may also be an external storage device of the computer device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the computer device. Of course, the memory may include both internal storage units and external storage devices of the computer device. In this embodiment, the memory is often used to store the operating system and various application software installed on the computer device, such as the program code of the automatic measurement method for ice-rock sorting degree in the rotating drum experiment based on vision recognition. In addition, the memory can also be used to temporarily store various types of data that have been output or will be output.
[0099] In some embodiments, the processor may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor is typically used to control the overall operation of the computer device. In this embodiment, the processor is used to run program code stored in the memory or process data, for example, to run the program code of the vision-based automatic measurement method for ice-rock sorting in a rotating drum experiment.
[0100] Fourthly, this application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the above-described automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition.
[0101] The computer-readable storage medium stores an interface display program that can be executed by at least one processor to cause the at least one processor to perform the steps of the above-described automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition.
[0102] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the automatic measurement method for ice and rock sorting degree based on visual recognition in the rotating drum experiment described in the embodiments of this application.
[0103] The invention will be further described below with reference to application examples.
[0104] Application examples In this application example, the outer diameter of the rotating drum 3 is 200 cm, the inner width of the opening of the rotating drum 3 is 30 cm, the material is smooth, scratch-resistant high-strength acrylic, and the rotation speed of the drum device is 10 rpm, that is, 10 revolutions per minute, to simulate the sorting behavior of the ice-rock mixture during low-speed rolling.
[0105] The materials used in the experiment were artificial ice particles 10 and natural sandstone particles, with a particle size controlled at 2 cm, and thoroughly mixed at a volume ratio of 1:1. To ensure the observability of the sorting process, the mixed particles were spread evenly to a thickness of approximately 20 cm at the center of the rotating drum 3 after being loaded, ensuring that all ice particles 10 were initially at the bottom. Camera 6 was equipped with a wide-angle lens to cover the entire top projection area of the rotating drum device. A flow velocity and depth sensor 8 was placed on one side of camera 6 to monitor the thickness and velocity of the debris flow at the lowest point of the rotating drum device. Simultaneously, a pore pressure and moisture content sensor was installed at the bottom of the rotating drum device to observe the pressure changes within the debris fluid during movement. To eliminate shadows and reflections in the images, ring-shaped LED cold light sources were placed on both sides of the rotating drum device to achieve uniform and stable lighting conditions.
[0106] Before the experiment began, the camera 6 and motor 4 were connected via a control computer to adjust the image viewing angle and focal length, and to manually calibrate the initial boundaries. Then, the drum device was turned on, and the image acquisition program was started. The image sampling frequency was set to 1 frame per second, and continuous acquisition lasted for 30 minutes, resulting in 1800 frames. All image data was fed into the system's preprocessing module in real time, where boundary cropping, grayscale conversion, and binarization were performed sequentially to maximize the extraction of particle boundaries and distribution features.
[0107] The preprocessed image enters the recognition and classification module. The system automatically identifies the pixel distribution of ice and rock particles in the image, calculates their areas (represented by the number of pixels), and outputs the sorting index 𝑆 for the current frame image. This index is defined as the proportion of the area of ice particles 10 in the total effective area, and the calculation formula is 𝑆 = A （ice） / (A （ice） +A （rock） ), where A （ice） With A （rock） The pixel areas of ice particle 10 and rock particle 9 in the image are shown respectively. The system outputs the sorting degree results in real time at a frequency of seconds. All data is recorded in CSV format, and a "time-sorting degree" change curve is automatically generated to analyze the stage characteristics of the sorting behavior.
[0108] Experimental results show that, Figure 4In the initial stage (0–3 minutes), ice particles 10 were located at the bottom of the rotating drum, with the surface entirely composed of rock particles 9. No obvious bright spots were observed in the images, and the sorting degree (K) was close to 0. As the drum continued to rotate, the less dense ice particles 10 gradually rose to the surface. From the 3rd minute onwards, exposed areas of ice particles 10 were visible in the images, and the sorting degree rapidly increased. From then until the 21st minute, ice particles 10 continued to accumulate at the top, and the sorting degree steadily increased from 0.3 to approximately 0.6, entering the accelerated sorting stage. From the 24th minute onwards, the system observed that the surface ice particles 10 tended to be saturated, with the sorting degree fluctuating above 0.85, eventually stabilizing, indicating that the material mixture had undergone sufficient sorting. The data extraction results were compared with those of the traditional method (which manually extracts the sorting degree from the side). The results show that the overall trend of the results obtained by the two methods is consistent, but the traditional extraction method can only select a limited number of characteristic time points for data extraction, resulting in a limited sampling frequency. The method of this invention provides high-precision sorting degree data across the entire time series, which can present more fluctuation data of the sorting process and is more conducive to detailed research on this physical process.
[0109] This application case fully verifies the stability, accuracy, and efficiency of the method of this invention in practical applications. Compared with traditional sidewall imaging methods, the top-view approach of this invention not only significantly improves image quality and avoids interference from scratches, water films, and wall reflections, but also provides stronger representativeness and data continuity, accurately capturing the physical evolution characteristics of ice-rock mixtures during the sorting process. The entire experimental process is highly automated, requiring no manual intervention, and exhibits good repeatability and data closed-loop characteristics. The constructed "image input – sorting recognition – index calculation – result export" process can also be flexibly applied to other types of binary mixture sorting experiments, such as sand-ice and snow-gravel particle systems, possessing broad promotional value and engineering adaptability.
[0110] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. An automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition, characterized in that, Includes the following steps: Image data acquisition; Image preprocessing; Identification of ice and rock particles in preprocessed images and pixel statistics of ice and rock particles; Sorting degree calculation, with real-time output of results; Specifically, the pixel area of the pixels identified as ice particles in the image is statistically analyzed and denoted as 𝐴. ice The pixel area of each pixel identified as a rock particle in the image is statistically analyzed and denoted as 𝐴. rock The total effective area of the image's field of view is denoted as 𝐴. total Let the sorting degree K be the proportion of ice particle area, and the calculation formula is as follows: 。 2. The automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition as described in claim 1, characterized in that, Image data acquisition includes the following steps: Install the rotary drum experimental system; Start the drum test system and perform real-time image acquisition.
3. The automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition as described in claim 2, characterized in that, The drum experimental system includes: A rotating drum device includes an annular rotating cylinder with an opening on its inner surface. A guard arm is fixed to the rotating cylinder and extends to the center point of the rotating cylinder. A support frame is located on both sides of the rotating drum device. A horizontal platform is provided on the support frame, and a motor is provided on the platform. The motor is connected to the guard arm at the center point of the rotating drum. A camera is located at the bottom of the platform and is aligned with the opening of the rotating drum.
4. The automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition as described in claim 3, characterized in that, At least two guard arms are provided, and multiple guard arms are evenly distributed on the rotating drum along the circumference of the rotating drum. Each guard arm extends to the center point of the rotating drum and converges to form a connecting part; and / or A flow velocity and depth sensor is installed at the bottom of the platform, and a pressure impact sensor is installed at the bottom of the rotating drum. The rotating drum experimental system also includes a control system, which is electrically connected to the flow velocity and depth sensor, the pressure impact sensor, and the motor, respectively; and / or At least one side of the rotating cylinder is made of transparent material.
5. The automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition as described in claim 4, characterized in that, Before the experiment begins, the camera is connected to the control system, and real-time image acquisition is initiated through the control program of the control system.
6. The automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition as described in claim 1, characterized in that, Image preprocessing includes performing boundary cropping, grayscale conversion, and binarization on each frame; and / or Pixel statistics include counting the number of pixels for ice particles and rock particles.
7. The automatic measurement method for ice and rock sorting degree in a rotating drum experiment based on visual recognition as described in claim 1, characterized in that, After preprocessing each frame of image, the sorting score of the current frame is output in real time. 𝑡 It also automatically generates a time-sorting curve.
8. An automatic measurement system for ice and rock sorting degree in a rotating drum experiment based on visual recognition, characterized in that, include: Image data acquisition module, used for image data acquisition; Image preprocessing module, used for image preprocessing; The recognition and pixel statistics module is used for the recognition of ice particles and rock particles in the preprocessed image, as well as the pixel statistics of ice particles and rock particles. The real-time output results module is used for sorting degree calculation and outputs the results in real time. Specifically, the pixel area of the pixels identified as ice particles in the image is statistically analyzed and denoted as 𝐴. ice The pixel area of each pixel identified as a rock particle in the image is statistically analyzed and denoted as 𝐴. rock The total effective area of the image's field of view is denoted as 𝐴. total Let the sorting degree K be the proportion of ice particle area, and the calculation formula is as follows: 。 9. A computer device, characterized in that, It includes a storage device and a processor, the storage device storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1-7.
10. A computer-readable storage medium, characterized in that, The device stores a computer program that, when executed by a processor, causes the processor to perform the steps of the method as described in any one of claims 1-7.