Particle motion image post-processing method based on MATLAB and PCC

Abstract

The invention relates to an image processing method, and aims to provide a particle motion image post-processing method based on MATLAB and PCC, which can realize rapid and accurate analysis of particle images and provide powerful support for performance analysis and optimization of a centrifugal pump. According to the technical scheme, the particle motion image post-processing method based on MATLAB and PCC is characterized by comprising the following steps: S1, image acquisition; s2, PCC pretreatment is carried out; s3, carrying out MATLAB (matrix laboratory) post-treatment; and S4, outputting and visualizing a result.

Description

Technical Field

[0001] The invention relates to an image processing method, in particular to a particle motion image post-processing method based on MATLAB and PCC. Background Art

[0002] As a widely used fluid conveying equipment, the performance and efficiency of centrifugal pumps directly affect the stability and economy of related industrial processes. When solid particles exist in the transport medium, the flow components of the centrifugal pump will be severely damaged, reducing the life and operating performance of the pump. Therefore, in the process of centrifugal pump performance analysis and optimization, it is very important to monitor and analyze the motion state of particles in the fluid in the pump. High-speed cameras can capture high-speed motion images of particles in the pump. In order to more accurately study the particle motion characteristics and explore the particle transport laws, the particle motion images obtained by the visualization experiment are processed. However, these images often contain a lot of noise and redundant information, and effective post-processing methods are needed to extract valuable information. Traditional image processing methods have many shortcomings in processing such complex particle motion images, such as slow processing speed, low precision, and complex operation. Therefore, it is urgent to propose a particle motion image post-processing method based on MATLAB and PCC to solve the above problems. Summary of the invention

[0003] The purpose of the present invention is to overcome the shortcomings of the above-mentioned background technology and provide a particle motion image post-processing method based on MATLAB and PCC, which should be able to achieve fast and accurate analysis of particle images and provide strong support for the performance analysis and optimization of centrifugal pumps.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A particle motion image post-processing method based on MATLAB and PCC, characterized in that it comprises the following steps:

[0006] S1: Image acquisition: high-speed image data of particle flow in a centrifugal pump is obtained by a high-speed camera device; the resolution and frame rate of the high-speed image data are 1280×800 and 8700fps respectively;

[0007] S2: PCC preprocessing: import the high-speed image data into the PCC software and preprocess the high-speed image data, including filtering and denoising, image enhancement, and preliminary segmentation steps, to improve image quality;

[0008] The image preprocessing also includes resizing and rotation correction processing of the image to further improve the accuracy and efficiency of image processing;

[0009] S3: MATLAB post-processing: The image data after PCC pre-processing is imported into MATLAB software for further processing, such as particle identification, feature extraction, and data analysis;

[0010] The further processing is: firstly, using the image processing algorithm in MATLAB to identify and track the particles, extracting the position, speed and size parameters of the particles and storing them in the data structure of MATLAB; finally, using the image processing toolbox of MATLAB to perform the particle identification, feature extraction and data analysis.

[0011] The particle identification and tracking includes a particle segmentation algorithm based on morphological processing, a particle tracking algorithm based on optical flow method, and a particle feature extraction and classification algorithm based on machine learning.

[0012] S4: Result output and visualization: The processing results are output in the form of images, data tables, etc. The identification results are visualized in the form of charts and reports using the drawing function of MATLAB, showing the movement trajectory and velocity distribution information of the particles, so as to conduct in-depth research and evaluation on the flow characteristics of the particles in the centrifugal pump;

[0013] The processing results include the quantity, type, and location information of the particles;

[0014] The visually displayed charts and reports include particle trajectory tracking diagrams, velocity vector diagrams, and statistical analysis results based on these data.

[0015] The beneficial effects of the present invention are: by combining the programming flexibility of MATLAB and the efficient computing power of the PCC platform, the processing precision and accuracy of high-speed flow particle images in a centrifugal pump are improved, the accurate capture and tracking of particle motion trajectories are achieved, and a visualization function is provided, which facilitates users to deeply analyze and understand the flow characteristics in the pump. The method has broad application value in the fields of fluid mechanics, materials science, and chemical engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of a particle motion image post-processing method based on MATLAB and PCC provided in an embodiment of the present invention.

[0017] Figure 2 Schematic diagram of a particle motion image shooting experimental platform according to an embodiment of the present invention.

[0018] Figure 3 It is the original image, edge detection and particle image recognition result in the centrifugal pump provided by the embodiment of the present invention.

[0019] Figure 4 It is the particle velocity vector calculation result provided by the embodiment of the present invention. DETAILED DESCRIPTION

[0020] The present invention will be described in detail below in conjunction with the embodiments shown in the accompanying drawings.

[0021] Example

[0022] Figure 2 In the particle motion image shooting experimental platform shown, the transparent centrifugal pump is a 5-blade transparent centrifugal pump. Under the condition of low particle volume fraction concentration, a high-speed photography experiment was carried out on the particles of POM material with a particle size of d = 3mm to obtain high-quality particle motion images. The particle velocity calculation program designed in MATLAB software based on Canny operator edge detection, Hough transform circle detection and nearest neighbor particle matching algorithm was used to analyze the distribution and movement law of particles in the centrifugal pump.

[0023] Reference Figure 1 The image post-processing method flow shown in the figure, a particle motion image post-processing method based on MATLAB and PCC provided by the present invention comprises the following steps:

[0024] (1) Image acquisition: Use a high-speed camera to capture the flow of particles inside the centrifugal pump in real time to obtain high-speed image data of particle movement. The high-speed image data must meet certain resolution and frame rate requirements, which are 1280×800 and 8700fps respectively, to ensure that the movement trajectory of the particles can be clearly captured.

[0025] (2) PCC preprocessing: The collected high-speed image data is transmitted to PCC in real time for preliminary processing. Before calculating the particle velocity of the image captured by the camera, the image collected by the high-speed camera needs to be preprocessed, including filtering and denoising, image enhancement, and preliminary segmentation steps. PCC preprocessing technology is optimized for the image characteristics in the centrifugal pump to improve image quality and processing efficiency.

[0026] Filtering and denoising: The built-in image processing algorithm of PCC is used to filter and denoise the image to reduce noise interference in the image and improve image quality.

[0027] Image enhancement: Using the real-time processing capability of PCC, the image is enhanced, such as contrast enhancement and brightness adjustment, to highlight the particle features and facilitate subsequent processing.

[0028] Preliminary segmentation: Based on the shape, size and other characteristics of the particles, the image is preliminarily segmented to separate the particles from the background, laying the foundation for subsequent fine processing.

[0029] (3) MATLAB post-processing: The image data after PCC pre-processing is imported into MATLAB software for further processing; particle identification, feature extraction, and data analysis are performed using MATLAB's powerful image processing toolbox;

[0030] Particle identification: Use image processing algorithms in MATLAB to perform more detailed image segmentation to ensure that each particle is accurately identified;

[0031] Feature extraction: Extract the shape, size, and motion trajectory characteristics of each particle to provide data support for subsequent analysis;

[0032] Data analysis: Based on the extracted feature information, the velocity, acceleration, and trajectory change state of the particles are statistically analyzed to evaluate the performance and efficiency of the centrifugal pump.

[0033] (4) Result output and visualization: The processing results are output in the form of images and data tables, including the number, type, and location information of the particles, for easy viewing and analysis by users; the recognition results are visualized in the form of charts and images using MATLAB’s drawing functions (such as imshow, plot, etc.), showing the particle motion trajectory and velocity distribution information (such as Figure 4 As shown in the figure), users can intuitively understand the movement state of particles in the centrifugal pump.

[0034] Preferably, in step (1), the particle motion image shooting experimental platform ( Figure 2As shown in the prior art), the particle motion image of the particles transported by the centrifugal pump was taken according to the experimental scheme and experimental steps. During the experiment, the particles were added or discharged from the solid-liquid two-phase circuit through the three-way valve, and the fluid could enter the water storage tank through valve 1 and valve 2, and the filter screen was designed at valve 1 and valve 2 to ensure that the particles could not flow into the water storage tank. The particle motion image acquisition process is: fill the water storage tank to more than two-thirds of the water, and ensure that the centrifugal pump is full of water. Open the exhaust hole to exhaust the gas in the pipeline system to ensure the smoothness of the fluid flow. Turn on the centrifugal pump and adjust the speed to achieve stable operation. Start the lighting and high-speed camera system, fine-tune the lighting conditions and camera lens, and ensure that the center of the impeller flow channel becomes a clear observation point. Enable the coordinate ruler in the high-speed camera software PCC, adjust the camera's viewing angle and position, so that the coordinate center is accurately aligned with the impeller center, and the vertical coordinate line is perfectly aligned with the edge of the pore. According to the experimental requirements, the particles of the specified volume fraction are accurately introduced into the system through the three-way valve. Allow the system to continue to cycle until the particles are evenly distributed in the two-phase circuit and the system is stable. Adjust the angle and distance of the LED light, and fine-tune the focus of the camera to obtain high-quality two-phase flow images. For each set of experiments, continuously capture and save at least 5,000 frames of particle flow images to ensure the adequacy and completeness of the data. Evaluate the quality of the captured images. If they do not meet the experimental standards, repeat the experiment and mark them to distinguish the data of different groups, and continue with the next set of experiments.

[0035] Preferably, in step (2), the captured particle motion image is stored as a video format (.mp4) file in the high-speed camera software PCC, and then imported into the MATLAB software, and the mp4 file is decomposed into a continuous image sequence according to the set image interpretation frame interval. The trajectory of the particles and the movement pattern of the particles before and after the collision can be obtained by superimposing the image sequence. The movement speed of the particles in the pump can be obtained by performing particle recognition and matching calculations on two consecutive images. Preliminary segmentation can use image segmentation methods to identify particles. Threshold segmentation and edge detection can both accurately identify the particles themselves from the image background, that is, segmenting the particles from the image background (such as Figure 3 shown).

[0036] Preferably, in step (3), the image transmission step uses the imread function or VideoReader class of MATLAB to import the image or image sequence containing particles. The specific steps of fine segmentation include: multi-threshold segmentation, region-based segmentation, morphological operation, and graph cut. For particles with different brightness or contrast, (multi-level application of Otsu method) is used to more accurately segment the image. Region-based segmentation techniques such as region growing, level set method or watershed algorithm are used to identify and separate particles. These methods can merge or segment regions based on pixel similarity (such as grayscale, color, texture, etc.). In the segmentation process, morphological operations (such as corrosion, expansion, opening operation, closing operation) are combined to remove small noise points, fill holes inside particles, or separate particles that are adhered to each other. For complex images, a graph cut algorithm is used, which models the image as a graph and finds the best segmentation boundary by minimizing an energy function. In the feature extraction step, the contours of the particles are first extracted using functions such as bwboundaries and bwtraceboundary, and contour smoothing or correction may be required to remove artifacts. The programs required in MATLAB are as follows:

[0037]

[0038]

[0039]

[0040]

[0041]

[0042] In brief, the particle motion image post-processing method based on MATLAB and PCC of the present invention comprises the following steps:

[0043] Step S1: Use a high-speed camera to photograph the inside of the centrifugal pump to obtain a high-speed camera image of particle movement.

[0044] Step S2: importing the high-speed image data into the PCC software for preprocessing; including: converting the color image into a grayscale image to reduce the amount of data processing; using conventional Gaussian filtering and median filtering methods to remove noise in the image and improve image quality; and using the histogram equalization method to enhance the image contrast and make the boundary between the particles and the background clearer;

[0045] Step S3: MATLAB post-processing: Use the multi-threshold segmentation algorithm in MATLAB (such as the Otsu method) to determine the optimal threshold, binarize the grayscale image, and separate the particles from the background; perform morphological operations such as corrosion, expansion, opening, and closing on the binary image to remove small noise points, fill holes inside the particles, and separate adherent particles; use the bwlabel function in MATLAB to mark the connected areas of the morphologically processed image and assign a unique label to each particle; extract the centroid, area, and shape feature information of each particle. Use a feature-based particle tracking algorithm (such as the KLT algorithm and particle filtering) to match and track particles in continuous image frames and record the trajectory of each particle;

[0046] Step S4: Result output and visualization

[0047] The basic information of the particles, such as coordinates, velocity, acceleration, and tracking trajectory data, are exported into files of a specific format (such as .csv, .mat, etc.) for subsequent processing and analysis.

[0048] Calculate and visualize the velocity field of particle motion, analyze velocity distribution and variation patterns. Perform statistical analysis on particle trajectories, extract trajectory characteristic parameters (such as average velocity, acceleration, trajectory length), and evaluate the stability and consistency of particle motion. Based on particle motion data, combined with the structural parameters and operating conditions of the centrifugal pump, use the fluid mechanics model to simulate the flow field, predict and explain the motion behavior of particles inside the centrifugal pump.

[0049] Output the post-processing results in the form of graphs, tables, reports, etc. for further analysis and use.

[0050] In summary, the above technical solution of the present invention realizes efficient and accurate post-processing of high-speed camera particle images in centrifugal pumps. This method not only improves the image quality, but also significantly improves the accuracy and efficiency of feature extraction and recognition, providing strong support for in-depth research on the flow characteristics inside centrifugal pumps. At the same time, this method has strong versatility and scalability, and can be widely used in image post-processing of other fluid machinery and flow systems.

[0051] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A particle motion image post-processing method based on MATLAB and PCC, characterized in that: The following steps are involved: S1: Image acquisition: high-speed image data of particle flow in the centrifugal pump is obtained by high-speed camera equipment; S2: PCC preprocessing: import the high-speed image data into the PCC software and preprocess the high-speed image data, including filtering and denoising, image enhancement, and preliminary segmentation steps, to improve image quality; S3: MATLAB post-processing: The image data after PCC pre-processing is imported into MATLAB software for further processing, such as particle identification, feature extraction, and data analysis; S4: Result output and visualization: The processing results are output in the form of images, data tables, etc. The identification results are visualized in the form of charts and reports using the drawing function of MATLAB, showing the movement trajectory and velocity distribution information of the particles, so as to conduct in-depth research and evaluation on the flow characteristics of the particles in the centrifugal pump.

2. The method according to claim 1, characterized in that: The resolution and frame rate of the high-speed image data in step S1 are 1280×800 and 8700 fps respectively.

3. The method according to claim 2, characterized in that: The further processing in step S3 is: firstly, the particles are identified and tracked using the image processing algorithm in MATLAB, and the position, speed and size parameters of the particles are extracted and stored in the data structure of MATLAB; finally, the particle identification, feature extraction and data analysis are performed using the image processing toolbox of MATLAB.

4. The method according to claim 3, characterized in that: The processing result in step S4 includes the quantity, type and position information of the particles.

5. The method according to claim 4, characterized in that: The PCC image preprocessing in step S2 also includes resizing and rotation correction processing of the image to further improve the accuracy and efficiency of image processing.

6. The method according to claim 5, characterized in that: The particle identification and tracking in step S3 includes a particle segmentation algorithm based on morphological processing, a particle tracking algorithm based on optical flow method, and a particle feature extraction and classification algorithm based on machine learning.

7. The method according to claim 6, characterized in that: The charts and reports visualized in step S4 include particle trajectory tracking diagrams, velocity vector diagrams, and statistical analysis results based on these data.

Images