Hydraulic equipment assembly regulation method and system based on image processing

Abstract

The application provides an image processing-based hydraulic equipment assembly regulation and control method and system, relates to the technical field of image processing, and comprises the following steps: analyzing a historical assembly vibration signal multidimensional domain feature, predicting a minimum vibration time point by an intelligent prediction model, acquiring a first historical collection image, introducing an image quality evaluation function to obtain a first image quality index, setting light, and performing assembly regulation and control of a first assembly link. The application can solve the technical problem of low assembly alignment accuracy caused by low quality of the collected real-time assembly image when the prior art regulates and controls the assembly of the hydraulic equipment based on the real-time assembly image. The application determines the optimal image collection time by analyzing the vibration signal during the assembly of the hydraulic equipment, optimizes the light setting during image collection, collects a high-quality real-time assembly image, finally realizes accurate regulation and control of high-precision hydraulic equipment assembly alignment, and achieves the technical effect of improving the assembly quality of the hydraulic equipment.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a hydraulic equipment assembly and control method and system based on image processing. Background Technology

[0002] With the continuous advancement of engineering machinery technology, hydraulic equipment, as a crucial component, is widely used in modern industry, involving precise assembly and control processes. During the assembly and control of hydraulic equipment, computer vision systems are often required for analysis and judgment. However, due to low image quality, such as low resolution and high noise, the computer analysis results are often inaccurate, thus affecting assembly precision.

[0003] In summary, existing technologies for assembling and controlling hydraulic equipment based on real-time assembly images suffer from low assembly alignment accuracy due to the poor quality of the acquired real-time assembly images. Summary of the Invention

[0004] The purpose of this application is to provide a hydraulic equipment assembly control method and system based on image processing, in order to solve the technical problem in the prior art that the low quality of the acquired real-time assembly images leads to low assembly alignment accuracy when hydraulic equipment is assembled and controlled based on real-time assembly images.

[0005] In view of the above problems, this application provides a hydraulic equipment assembly and control method and system based on image processing.

[0006] In a first aspect, this application provides a hydraulic equipment assembly control method based on image processing. The method is implemented through a hydraulic equipment assembly control system based on image processing. The method includes: acquiring historical assembly vibration signals, where the historical assembly vibration signals refer to the vibration signals of the hydraulic equipment during a first assembly stage in a historical assembly process; performing multi-dimensional domain feature analysis on the historical assembly vibration signals to obtain vibration signal domain feature information; calling an intelligent prediction model to predict the minimum vibration time point of the vibration signal domain feature information to obtain a target time point; and acquiring historical assembly light parameters, where the historical assembly light parameters refer to the light parameters of the hydraulic equipment during the first assembly stage in a historical assembly process. Extract the first historical lighting parameter from the historical assembly lighting parameters and obtain the first historical acquisition image under the first historical lighting parameter; introduce an image quality evaluation function to obtain the first image quality index of the first historical acquisition image, and when the first image quality index reaches the quality index limit, use the first historical lighting parameter as the preset lighting parameter; set the lighting for the first assembly stage of the hydraulic equipment based on the preset lighting parameter, and start the image acquisition device at the target time point; acquire assembly images of the first assembly stage through the image acquisition device to obtain the first assembly image, and adjust the assembly of the first assembly stage according to the first assembly image.

[0007] Secondly, this application also provides an image processing-based hydraulic equipment assembly and control system for executing the image processing-based hydraulic equipment assembly and control method as described in the first aspect. The system includes: a historical assembly vibration signal acquisition module, which acquires historical assembly vibration signals, referring to the vibration signals of the hydraulic equipment during the first assembly stage in a historical assembly process; a multi-dimensional domain feature analysis module, which performs multi-dimensional domain feature analysis on the historical assembly vibration signals to obtain vibration signal domain feature information; an intelligent prediction model invocation module, which invokes an intelligent prediction model to predict the minimum vibration time point of the vibration signal domain feature information to obtain a target time point; and a historical assembly light parameter acquisition module, which acquires historical assembly light parameters, referring to the vibration signals of the hydraulic equipment during a historical assembly process. The system includes: a light parameter extraction module for the first assembly stage; a light parameter extraction module for extracting a first historical light parameter from the historical assembly light parameters and acquiring a first historical acquisition image under the first historical light parameter; an image quality evaluation module for introducing an image quality evaluation function to obtain a first image quality index of the first historical acquisition image, and using the first historical light parameter as a preset light parameter when the first image quality index reaches a quality index limit; a light setting module for setting the light for the first assembly stage of the hydraulic equipment based on the preset light parameter, and starting the image acquisition device at the target time point; and an image acquisition and assembly control module for acquiring an assembly image of the first assembly stage through the image acquisition device to obtain a first assembly image, and controlling the assembly of the first assembly stage based on the first assembly image.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] By acquiring historical assembly vibration signals, which refer to the vibration signals of the first assembly stage of the hydraulic equipment during historical assembly; performing multi-dimensional domain feature analysis on the historical assembly vibration signals to obtain vibration signal domain feature information; calling an intelligent prediction model to predict the minimum vibration time point of the vibration signal domain feature information to obtain a target time point; acquiring historical assembly lighting parameters, which refer to the lighting parameters of the first assembly stage of the hydraulic equipment during historical assembly; extracting the first historical lighting parameter from the historical assembly lighting parameters and acquiring the first historical acquisition image under the first historical lighting parameter; introducing an image quality evaluation function to obtain the first image quality index of the first historical acquisition image, and when the first image quality index reaches the quality index limit, using the first historical lighting parameter as the preset lighting parameter; setting the lighting for the first assembly stage of the hydraulic equipment based on the preset lighting parameter, and starting the image acquisition device at the target time point; acquiring assembly images of the first assembly stage through the image acquisition device to obtain the first assembly image, and adjusting the assembly of the first assembly stage based on the first assembly image. In other words, by analyzing the vibration signals during the assembly of hydraulic equipment to determine the optimal image acquisition time, and simultaneously analyzing and optimizing the lighting settings during image acquisition, high-quality real-time assembly images can be acquired. Ultimately, this achieves precise control of the alignment of high-precision hydraulic equipment assembly, thereby improving the technical effect of hydraulic equipment assembly quality.

[0010] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0012] Figure 1 This is a schematic flowchart of the hydraulic equipment assembly and control method based on image processing in this application;

[0013] Figure 2This is a schematic diagram of the hydraulic equipment assembly and control system based on image processing, as described in this application.

[0014] Figure labeling: Historical assembly vibration signal acquisition module 11, multi-dimensional domain feature analysis module 12, intelligent prediction model calling module 13, historical assembly light parameter acquisition module 14, light parameter extraction module 15, image quality assessment module 16, light setting module 17, image acquisition and assembly control module 18. Detailed Implementation

[0015] This application provides a hydraulic equipment assembly control method and system based on image processing, solving the technical problem in existing technologies where the low quality of real-time assembly images leads to low assembly alignment accuracy when controlling hydraulic equipment assembly based on real-time assembly images. By analyzing the vibration signals during hydraulic equipment assembly to determine the optimal image acquisition time and simultaneously analyzing and optimizing the lighting settings during image acquisition, high-quality real-time assembly images are acquired, ultimately achieving precise control of high-precision hydraulic equipment assembly alignment and improving the technical effect of hydraulic equipment assembly quality.

[0016] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0017] Example 1

[0018] Please see the appendix Figure 1 This application provides a hydraulic equipment assembly and control method based on image processing, wherein the method is applied to a hydraulic equipment assembly and control system based on image processing, and the method specifically includes the following steps:

[0019] Step 1: Obtain historical assembly vibration signals, which refer to the vibration signals of the hydraulic equipment during the first assembly stage in the historical assembly process.

[0020] Specifically, this involves collecting vibration data from past assembly processes of hydraulic equipment, particularly during the first assembly stage. The first assembly stage refers to the first critical phase in the equipment assembly process, involving the installation of the first major component or part into the equipment, or initiating the first critical operation in the assembly process. These vibration signals are electrical signals converted from vibrations generated during equipment operation or assembly. They are typically acquired through vibration sensors installed on the equipment, which capture the mechanical vibrations generated during operation and contain rich information about the equipment's operating status, such as vibration amplitude, frequency, and phase. This information can reveal the dynamic behavior inside the equipment, such as wear, imbalance, loosening, or other mechanical faults in components. By acquiring historical assembly vibration signals, potential anomalies during the assembly process can be identified, allowing for the early detection of potential faults.

[0021] Step 2: Perform multi-dimensional domain feature analysis on the historical assembly vibration signal to obtain vibration signal domain feature information.

[0022] Specifically, the collected historical assembly vibration signals undergo preprocessing, including denoising, filtering, and normalization. This is primarily to eliminate random noise and irrelevant information in the signals, ensuring the accuracy of the analysis results. Denoising helps eliminate noise introduced by environmental factors or sensor errors, filtering removes unwanted frequency components from the signal, and normalization unifies the amplitude range between different signals. Multi-dimensional domain feature analysis is then performed on the preprocessed vibration signals, processing them in multiple dimensions such as the time domain, frequency domain, and time-frequency domain to extract domain feature information. This information reflects the equipment's operating status, potential faults, and performance indicators. Multi-dimensional domain feature analysis is a data analysis method that examines datasets from multiple perspectives to reveal hidden patterns, relationships, or features. It is frequently used in big data analysis, machine learning, image processing, and speech recognition. In multi-dimensional domain feature analysis, data is not only analyzed in terms of its characteristics in a single dimension but also the interactions and correlations between different dimensions are examined. The analysis includes several domain-specific approaches: time-domain analysis examines the waveform, amplitude, and trend of vibration signals, including calculating statistical parameters such as average value, peak value, and root mean square value; frequency-domain analysis uses methods like Fourier transform to convert the vibration signal from the time domain to the frequency domain, analyzing its frequency components such as spectrum, dominant frequency, and sideband distribution; and time-frequency analysis employs methods like short-time Fourier transform or wavelet transform to analyze the frequency components of the signal at different time points, identifying the signal's instantaneous characteristics. These analyses extract domain-specific features of the vibration signal, reflecting the equipment's operating status, potential faults, and performance indicators. This multi-dimensional domain feature analysis, utilizing techniques from different domains to gain a deeper understanding of vibration signals, is crucial for training intelligent prediction models and assembly control, improving assembly accuracy and efficiency.

[0023] Step 3: Call the intelligent prediction model to predict the minimum vibration time point of the vibration signal domain feature information to obtain the target time point.

[0024] Specifically, new vibration signals are input into a trained intelligent prediction model. The model then predicts the minimum vibration time point, or target time point, based on learned patterns. The minimum vibration time point refers to the specific moment when the vibration amplitude or intensity is at its lowest during equipment operation, allowing us to understand when the equipment will reach its optimal operating state. In the assembly process of hydraulic equipment, the minimum vibration time point may correspond to the installation or adjustment step of a critical component. The target time point refers to a future point in time; the prediction model determines the minimum vibration time point based on historical vibration signal data and multi-dimensional domain feature analysis. By utilizing the intelligent prediction model, a deeper understanding of vibration signals can be achieved, providing accurate predictions for equipment operating status assessment and fault diagnosis.

[0025] Step 4: Obtain historical assembly lighting parameters, which refer to the lighting parameters of the first assembly stage of the hydraulic equipment during historical assembly.

[0026] Specifically, lighting conditions are crucial for obtaining high-quality assembly images during the assembly process of hydraulic equipment. Historical assembly processes record and analyze the ambient lighting conditions of the first assembly stage, including parameters such as light source type, brightness, color temperature, direction of illumination, distance, and uniformity of illumination. Lighting parameters specifically refer to various light characteristics in the assembly environment that affect visual perception and equipment performance. These parameters collectively determine the distribution and intensity of light, thus affecting image quality and analyzability. By analyzing historical data, we can identify which lighting parameters produce high-quality assembly images, allowing us to use similar lighting conditions in future assembly processes. Understanding and utilizing lighting parameters helps identify and resolve problems during assembly, such as insufficient or excessive lighting, shadows, and glare, thereby further improving the quality of the acquired real-time assembly images.

[0027] Step 5: Extract the first historical ray parameter from the historical assembly ray parameters, and obtain the first historical acquisition image under the first historical ray parameter.

[0028] Specifically, the system selects and extracts lighting parameters used in the first assembly stage from a database of recorded historical assembly lighting parameters. These parameters include the type of light source, light intensity, color temperature, illumination direction, distance, and illumination uniformity. Based on these extracted historical lighting parameters, the system acquires first historical images under those lighting conditions to understand and evaluate the impact of the lighting conditions used in the first assembly stage on image quality. Acquired images refer to those captured by monitoring cameras or specialized image acquisition equipment during equipment assembly, used to record assembly details, operational processes, or to inspect assembly quality. These images are acquired to analyze assembly results related to specific lighting conditions or for subsequent image processing and analysis, such as object recognition and defect detection. By analyzing and utilizing these parameters and images, optimal lighting conditions can be employed in future assembly processes to obtain high-quality assembly images.

[0029] Step 6: Introduce an image quality evaluation function to obtain the first image quality index of the first historically acquired image, and when the first image quality index reaches the quality index limit, use the first historical lighting parameters as preset lighting parameters.

[0030] Specifically, an image quality assessment function is introduced to quantify the quality of the first historically acquired image. This function is typically based on a series of image quality metrics, such as sharpness, contrast, noise level, and edge information, to determine whether the image meets predetermined quality standards. A comprehensive image quality index is obtained by calculating these metrics, reflecting the overall quality level of the image. When the first image quality index reaches a quality index limit, it indicates that the image quality has met the preset standard. This limit is a threshold set based on actual needs and experience, defining the minimum acceptable image quality level. When the first image quality index reaches this limit, it means the image quality meets the requirements and can be used for subsequent analysis and processing. The first historical lighting parameters are used as preset lighting parameters, representing the optimal choice for obtaining high-quality images under current conditions, ensuring high-quality assembly images are obtained under similar conditions in the future. Through image quality assessment, it is ensured that the images acquired during the assembly process meet the quality standards.

[0031] Step 7: Set the lighting for the first assembly stage of the hydraulic equipment based on the preset lighting parameters, and start the image acquisition device at the target time point.

[0032] Specifically, the system uses a first image quality index obtained from the first historically acquired image through an image quality evaluation function, and preset lighting parameters determined by other relevant vibration signal feature information obtained through multi-dimensional domain feature analysis. Based on these preset lighting parameters, lighting settings are configured for the first assembly stage of the hydraulic equipment, including adjusting the position, angle, and brightness of the light source to ensure that the lighting conditions within the assembly area meet preset standards, reducing the impact of lighting on image quality. Secondly, the image acquisition device is activated at the optimal time for critical operations, specifically the minimum vibration time point predicted by the intelligent prediction model. Activating the image acquisition device at this time ensures image acquisition when equipment vibration is minimal, capturing image information of the equipment in its optimal operating state or at critical moments, resulting in higher image quality. The image acquisition device, activated at the target time point, is used to acquire images of the first assembly stage of the hydraulic equipment. It may be a camera, sensor, or other image acquisition tool capable of capturing image information during the assembly process. By optimizing lighting conditions and image acquisition timing, the accuracy and efficiency of assembly can be improved.

[0033] Step 8: Acquire assembly images of the first assembly stage using the image acquisition device to obtain a first assembly image, and adjust the assembly of the first assembly stage based on the first assembly image.

[0034] Specifically, devices used to capture image information during the assembly process, such as high-resolution cameras, industrial cameras, or other vision sensors capable of capturing details, capture images of the assembly process in real time and provide detailed information. During assembly, these devices automatically or manually capture images of the first assembly stage according to preset parameters and timings. The resulting first assembly image records visual details of the equipment, components, or operations in the assembly stage, reflecting the actual situation during assembly, including the assembly status of parts, assembly sequence, and assembly quality. Based on the first assembly image, the first assembly stage can be controlled, such as adjusting the assembly sequence, optimizing the assembly process, and improving assembly tools. Assembly control uses the information provided by the images to guide and adjust the assembly process. If the image shows that a component is not in the correct position, adjustments can be made immediately to ensure assembly quality. Alternatively, if the image reveals an unexpected design problem, the assembly process needs to be modified or the assembly guidelines updated. By analyzing the assembly information in the images, problems in the assembly process can be identified and resolved in a timely manner, improving the visualization and controllability of the assembly process.

[0035] Furthermore, this application also includes the following steps:

[0036] A training vibration signal is acquired, and its spectrum is obtained based on the Fast Fourier Transform principle. Spectral analysis is performed on the spectrum to obtain the training spectral density, and the training target time point range corresponding to the peak value of the training signal in the training spectral density is back-matched. This training target time point range is embedded into a first time point predictor. The training target time point of the training vibration signal is obtained, where the training target time point refers to the time point when the vibration is at its minimum. Supervised machine learning is performed on the training dataset constructed based on the vibration signal domain feature information and the training target time point to obtain a second time point predictor. The first time point predictor and the second time point predictor together form the intelligent prediction model.

[0037] Specifically, historical vibration signal data is collected, namely, actual vibration records of hydraulic equipment during the assembly process, including equipment operating status and assembly quality. The vibration signal in the time domain is converted to the frequency domain using a Fast Fourier Transform (FFT) to obtain the spectrum, showing the energy distribution of the vibration signal at different frequencies. Spectral analysis is performed on the spectrum to obtain the spectral density, i.e., the energy density of the vibration signal at different frequencies. Through spectral analysis, peaks in the signal are identified. These peaks typically correspond to specific states or events during equipment operation, such as the equipment's resonant frequency or the frequency characteristics of specific operating steps. Through reverse matching, the frequency components corresponding to these peaks can be mapped back to the original time domain, thus obtaining the time point range corresponding to the peaks of the training signal. The training target time point information identified through spectral analysis and other steps is integrated into the first time point predictor, i.e., the model used to predict the future state of the vibration signal, meaning that this key time point information is included as part of the prediction model.

[0038] Training vibration signals refer to vibration signal data recorded during the historical assembly process of hydraulic equipment. The goal is to extract the time point with the minimum vibration amplitude from this data. During the assembly process, this time point corresponds to the internal equilibrium state of the equipment or the optimal time for critical operations. Vibration signal domain feature information extracted from the vibration signals includes frequency components, amplitude variations, and time-domain statistical parameters. This vibration signal domain feature information is combined with the target training time point to form a training dataset. This dataset contains the input (feature information) and output (target time point) for training the machine learning model. Supervised machine learning algorithms, such as linear regression, support vector machines, and neural networks, are used to learn from the training dataset. Through training, the machine learning model can predict key time points in the vibration signals, training a second time-point predictor. This predictor, based on historical data, can predict future time points based on new vibration signal feature information. Combining the first predictor based on a statistical model and the second predictor based on machine learning forms an intelligent prediction model.

[0039] The basic principle of the Fast Fourier Transform (FFT) is that any periodic signal can be represented as a superposition of sine and cosine waves. FFT decomposes these sine and cosine waves into their frequency components and calculates the amplitude and phase of each frequency component. This yields the signal's spectrum, showing the energy distribution at different frequencies. It can convert complex time-domain vibration signals into a spectrum, thus revealing the signal's frequency components more clearly. Especially in the assembly process of hydraulic equipment, analyzing the spectrum of vibration signals can identify internal fault modes, optimize the assembly process, and improve assembly quality. Spectral analysis is a detailed study of the vibration signal spectrum, aiming to understand the energy distribution of the signal at different frequencies. Through spectral analysis, the main frequency components, frequency distribution patterns, and any abnormal frequency components of the signal can be identified. Utilizing historical data and advanced machine learning techniques, the minimum vibration time point of the vibration signal can be accurately predicted, thereby optimizing the assembly process.

[0040] Furthermore, this application also includes the following steps:

[0041] The training vibration signal is subjected to time-domain feature acquisition to obtain the training signal time-domain features; the training vibration signal spectrum is subjected to frequency-domain feature acquisition to obtain the training signal frequency-domain features; the training signal time-domain features and the training signal frequency-domain features constitute the vibration signal domain feature information.

[0042] Specifically, in the assembly process of hydraulic equipment, to optimize the assembly process and improve assembly quality, in-depth analysis of vibration signals is conducted, including time-domain feature acquisition and frequency-domain feature acquisition, which together constitute the domain characteristic information of the vibration signal. Time-domain feature acquisition refers to the analysis and processing of vibration signals in the time domain to extract parameters reflecting signal characteristics. This typically includes calculating the statistical characteristics of the signal, such as mean, variance, peak value, and root mean square (RMS) value. These parameters describe the signal's intensity and variation trend, helping to understand the signal's dynamic behavior. Among these, the peak value and RMS value reflect the signal's amplitude and energy level, while the variance describes the signal's fluctuation degree. Frequency-domain feature acquisition refers to the analysis and processing of vibration signals in the frequency domain to extract information reflecting the signal's frequency components. This is usually achieved through Fast Fourier Transform (FFT), converting the time-domain signal into a frequency-domain signal. Frequency-domain features include frequency, amplitude, and phase, revealing the signal's frequency components and energy distribution. Among these, the dominant frequency and side frequencies reflect the equipment's main vibration modes and fault characteristics. Combining the time-domain and frequency-domain features of the training signal to form vibration signal domain feature information provides a comprehensive understanding of the characteristics of the vibration signal, which can improve the accuracy and robustness of the prediction model.

[0043] Furthermore, step six of this application includes:

[0044] ;in, This refers to the first image quality index. This refers to the first image feature value of the first historically acquired image. This refers to the quality feedback coefficient. This refers to the coordinates in the first historically acquired image. DC coefficient of the cut image patch, This refers to the coordinates in the first historically acquired image. The AC coefficient of the segmented image patch, This refers to the influence factor of the DC coefficient on the feature value of the first image. This refers to the influence factor of the AC coefficient on the first image feature value, and .

[0045] Specifically, in order to quantify the quality of the first historically acquired image, an image quality evaluation function is introduced, the expression of which is as follows:

[0046] ;in, This refers to the first image quality index, which is a comprehensive indicator used to quantify the quality of an image. It evaluates the overall quality level of an image by comprehensively considering multiple features and parameters of the image, and determines whether it has met the preset quality standards. It refers to the first image feature value of the first historically acquired image. This value is calculated based on the DC coefficient and AC coefficient of each pixel in the image and is used to evaluate the features and quality of the image. This refers to the quality feedback coefficient, which is a weighting coefficient used to adjust the influence of the first image feature value on the final image quality index. The contribution of different features to image quality assessment can be adjusted according to needs. This refers to the coordinates in the first historically acquired image. The DC coefficient of a cut image patch refers to the low-frequency components of the image, which usually reflects the background or average brightness of the image and is very important for evaluating the sharpness and contrast of the image. This refers to the coordinates in the first historically acquired image. The AC coefficient of a segmented image patch refers to the high-frequency components of the image. It is the deviation of the pixel value in the image patch from the average value. It usually reflects the detail and edge information of the image and is very important for evaluating the resolution and detail sharpness of the image. It refers to the influence factor of the DC coefficient on the first image feature value. It is a parameter used to adjust the influence of the DC coefficient in the calculation of the first image feature value. It determines the degree of contribution of the DC coefficient to the final image quality assessment. The influence factor is multiplied by the DC coefficient to represent the influence of the DC component on the image quality. It refers to the influence factor of the AC coefficient on the first image feature value. It is a parameter used to adjust the influence of the AC coefficient in the calculation of the first image feature value and determines the degree of contribution of the AC coefficient to the final image quality assessment. These two factors jointly determine the calculation of image feature values, and their relative importance is fixed, ensuring that the image feature values ​​are balanced when considering DC and AC components. By adjusting the influencing factors, the impact of low-frequency and high-frequency components on image feature values ​​can be balanced. The image quality evaluation function assesses image quality by calculating the low-frequency and high-frequency components of the image and their degree of influence on image feature values. When the image quality meets the preset standard, appropriate lighting parameters can be adopted to optimize the image acquisition process and improve assembly quality and efficiency.

[0047] Furthermore, step six of this application includes:

[0048] The first historical image is cut to obtain a first cutting result, which includes multiple cut image blocks with coordinate identifiers; a first cut image block is extracted from the multiple cut image blocks with coordinate identifiers, and the first cut image block corresponds to a first coordinate; the first discrete cosine transform coefficient of the first cut image block is obtained, which includes a first DC coefficient and a first AC coefficient; a first mapping relationship is established between the first coordinate and the first DC coefficient and the first AC coefficient; and the image quality evaluation function calculates the first image quality index based on the first mapping relationship.

[0049] Specifically, the first historically acquired image is segmented into multiple small image blocks, each becoming a segmented image block. These segmented image blocks have the same size and are uniformly distributed throughout the image. Each image block contains local information of the image and is marked with its position (coordinate identifier) ​​in the original image. Image blocks with specific coordinate identifiers are selected; the first segmented image block is the image block at a specified coordinate, and the corresponding first coordinate refers to the position coordinates of the first segmented image block in the original image. A set of coefficients is obtained by performing a Discrete Cosine Transform (DCT) on the first segmented image block, including a first DC coefficient and a first AC coefficient. The DC coefficient reflects the average brightness of the image block, while the AC coefficient reflects the texture and detail information of the image block. A correlation is established between the coordinates of the first segmented image block and the DCT coefficients, allowing the system to find and utilize the corresponding DCT coefficients based on the coordinates of the image block for image quality assessment. The image quality assessment function calculates the quality index of each segmented image block, and these indices are combined using the established mapping relationship to obtain the overall image quality index.

[0050] Discrete Cosine Transform (DCT) is a commonly used image processing technique that transforms image patches from the pixel domain to the frequency domain. The first image patch is a local region in the original image, typically a fixed-size rectangular block, such as 8x8 or 16x16 pixels. Applying DCT to this patch yields a DCT coefficient matrix of the same size. In this matrix, the DC coefficients (top-left coefficients) represent the average brightness of the image patch, which is the average of all pixel values. The AC coefficients are all other coefficients except for the DC coefficients. These coefficients correspond to non-zero frequency components in the frequency domain, where low-frequency AC coefficients describe large-scale texture, while high-frequency AC coefficients describe small-scale details and edge information. By segmenting and analyzing the image, image quality can be assessed more accurately.

[0051] Furthermore, step six of this application includes:

[0052] The first saliency map of the first historically acquired image is obtained based on the Itti-Koch visual saliency principle; the first image block overlap degree between the target saliency image block and the predetermined saliency image block in the first saliency map is obtained, wherein the target saliency image block is the set of pixels in the first saliency map whose saliency value meets the predetermined limit; the first image quality index is adjusted using the first image block overlap degree as an adjustment coefficient.

[0053] Specifically, the Itti-Koch visual saliency principle is applied to analyze the first historically acquired image. By calculating the saliency values ​​of different regions in the image, regions with high saliency are identified, resulting in a first saliency map. A saliency threshold is pre-set; only pixels whose saliency values ​​exceed this threshold are considered saliency. Pixel sets whose saliency values ​​meet the pre-set threshold are identified from the first saliency map; these pixels have high saliency in the image and are called target saliency image patches. The degree of overlap between the target saliency image patches in the first saliency map and the pre-set saliency image patches is calculated. By calculating the proportion of pixels with the same value in these two image patches, the overlap degree is obtained, which measures the degree of matching between the target saliency image patch and the predefined saliency standard. The overlap degree of the first image patch is used as an adjustment coefficient to adjust the value of the first image quality index. If the overlap degree between the target saliency image patch and the pre-set saliency image patch is high, the target saliency image patch and the pre-set saliency image patch are very well matched, increasing the value of the first image quality index, indicating higher image quality. The Itti-Koch visual saliency principle is a method for computational vision systems to detect salient regions in an image. It identifies highly salient areas by calculating the saliency values ​​of different regions within the image. Salientity values ​​are typically related to the contrast of features such as color, brightness, and texture, as well as the spatial arrangement of these features. Overlap is calculated by comparing the number of overlapping pixels between two image patches. For example, if a target salient image patch overlaps with a predetermined salient image patch by 80% of its pixels, then the overlap is 0.8. Combining the Itti-Koch visual saliency principle with image quality assessment helps improve the accuracy and reliability of image quality evaluation.

[0054] Furthermore, this application also includes the following steps:

[0055] Read the predetermined light characteristics and obtain the first historical light parameters based on the predetermined light characteristics, wherein the predetermined light characteristics include light source quality, light source intensity and light interference.

[0056] Specifically, the system accesses a database storing preset fiber optic characteristic data, which may include parameters such as light source type, light source quality, light source intensity, and light interference. After determining the required first historical ray parameters, the system retrieves the corresponding ray parameter records from the database based on the specific values ​​of these parameters, such as specific light source quality and intensity. Light source quality, light source intensity, and light interference are key factors affecting lighting effects and image acquisition quality. Light source quality includes brightness, color temperature, and spectral distribution, directly impacting lighting effects. Light source intensity refers to the brightness level of the light source; higher intensity results in better lighting, but excessively strong light sources may produce shadows or glare, affecting image quality. Light interference refers to the degree of influence of the light source on the surrounding environment, including light reflection, refraction, and scattering. Higher light interference may result in more noise and artifacts in the image, affecting image clarity and detail. By reading and analyzing the predetermined ray characteristics, the first historical ray parameters can be obtained. During image acquisition, the brightness, color temperature, and other parameters of the light source can be adjusted based on the first historical ray parameters to optimize image quality.

[0057] In summary, the image processing-based hydraulic equipment assembly and control method provided in this application has the following technical effects:

[0058] By acquiring historical assembly vibration signals, which refer to the vibration signals of the first assembly stage of the hydraulic equipment during historical assembly; performing multi-dimensional domain feature analysis on the historical assembly vibration signals to obtain vibration signal domain feature information; calling an intelligent prediction model to predict the minimum vibration time point of the vibration signal domain feature information to obtain a target time point; acquiring historical assembly lighting parameters, which refer to the lighting parameters of the first assembly stage of the hydraulic equipment during historical assembly; extracting the first historical lighting parameter from the historical assembly lighting parameters and acquiring the first historical acquisition image under the first historical lighting parameter; introducing an image quality evaluation function to obtain the first image quality index of the first historical acquisition image, and when the first image quality index reaches the quality index limit, using the first historical lighting parameter as the preset lighting parameter; setting the lighting for the first assembly stage of the hydraulic equipment based on the preset lighting parameter, and starting the image acquisition device at the target time point; acquiring assembly images of the first assembly stage through the image acquisition device to obtain the first assembly image, and adjusting the assembly of the first assembly stage based on the first assembly image. In other words, by analyzing the vibration signals during the assembly of hydraulic equipment to determine the optimal image acquisition time, and simultaneously analyzing and optimizing the lighting settings during image acquisition, high-quality real-time assembly images can be acquired. Ultimately, this achieves precise control of the alignment of high-precision hydraulic equipment assembly, thereby improving the technical quality of hydraulic equipment assembly.

[0059] Example 2

[0060] Based on the image processing-based hydraulic equipment assembly and control method described in the foregoing embodiments, and using the same inventive concept, this application also provides an image processing-based hydraulic equipment assembly and control system. Please refer to the appendix. Figure 2 The system includes:

[0061] Historical assembly vibration signal acquisition module 11 is used to acquire historical assembly vibration signals, which refer to the vibration signals of the hydraulic equipment in the first assembly stage during the historical assembly process.

[0062] The multi-dimensional domain feature analysis module 12 is used to perform multi-dimensional domain feature analysis on the historical assembly vibration signal to obtain vibration signal domain feature information.

[0063] The intelligent prediction model invocation module 13 is used to invoke the intelligent prediction model to predict the minimum vibration time point of the vibration signal domain feature information and obtain the target time point.

[0064] Historical assembly light parameter acquisition module 14 is used to acquire historical assembly light parameters, which refer to the light parameters of the first assembly stage of the hydraulic equipment during historical assembly.

[0065] The light parameter extraction module 15 is used to extract the first historical light parameter from the historical assembly light parameters and obtain the first historical acquisition image under the first historical light parameter.

[0066] Image quality assessment module 16 is used to introduce an image quality assessment function to obtain a first image quality index of the first historically acquired image, and when the first image quality index reaches the quality index limit, the first historical lighting parameters are used as preset lighting parameters.

[0067] The light setting module 17 is used to set the light for the first assembly stage of the hydraulic equipment based on the preset light parameters, and to start the image acquisition device at the target time point.

[0068] The image acquisition and assembly control module 18 is used to acquire assembly images of the first assembly stage through the image acquisition device to obtain a first assembly image, and to control the assembly of the first assembly stage based on the first assembly image.

[0069] Furthermore, the multi-dimensional domain feature analysis module 12 in the system is also used for:

[0070] A training vibration signal is acquired, and its spectrum is obtained based on the Fast Fourier Transform principle. Spectral analysis is performed on the spectrum to obtain the training spectral density, and the training target time point range corresponding to the peak values ​​of the training signal in the training spectral density is back-matched. This training target time point range is embedded into a first time point predictor. The training target time point of the training vibration signal is obtained, where the training target time point refers to the time point when the vibration is at its minimum. Supervised machine learning is performed on the training dataset constructed based on the vibration signal domain feature information and the training target time point to obtain a second time point predictor. The first time point predictor and the second time point predictor together form the intelligent prediction model.

[0071] Furthermore, the multi-dimensional domain feature analysis module 12 in the system is also used for:

[0072] The training vibration signal is subjected to time-domain feature acquisition to obtain the training signal time-domain features; the training vibration signal spectrum is subjected to frequency-domain feature acquisition to obtain the training signal frequency-domain features; the training signal time-domain features and the training signal frequency-domain features constitute the vibration signal domain feature information.

[0073] Furthermore, the light parameter extraction module 15 in the system is also used for:

[0074] Read the predetermined light characteristics and obtain the first historical light parameters based on the predetermined light characteristics, wherein the predetermined light characteristics include light source quality, light source intensity and light interference.

[0075] Furthermore, the image quality assessment module 16 in the system is also used for:

[0076] ;in, This refers to the first image quality index. This refers to the first image feature value of the first historically acquired image. This refers to the quality feedback coefficient. This refers to the coordinates in the first historically acquired image. DC coefficient of the cut image patch, This refers to the coordinates in the first historically acquired image. The AC coefficient of the segmented image patch, This refers to the influence factor of the DC coefficient on the feature value of the first image. This refers to the influence factor of the AC coefficient on the first image feature value, and .

[0077] Furthermore, the image quality assessment module 16 in the system is also used for:

[0078] The first historical image is cut to obtain a first cutting result, which includes multiple cut image blocks with coordinate identifiers; a first cut image block is extracted from the multiple cut image blocks with coordinate identifiers, and the first cut image block corresponds to a first coordinate; the first discrete cosine transform coefficient of the first cut image block is obtained, which includes a first DC coefficient and a first AC coefficient; a first mapping relationship is established between the first coordinate and the first DC coefficient and the first AC coefficient; and the image quality evaluation function calculates the first image quality index based on the first mapping relationship.

[0079] Furthermore, the image quality assessment module 16 in the system is also used for:

[0080] The first saliency map of the first historically acquired image is obtained based on the Itti-Koch visual saliency principle; the first image block overlap degree between the target saliency image block and the predetermined saliency image block in the first saliency map is obtained, wherein the target saliency image block is the set of pixels in the first saliency map whose saliency value meets the predetermined limit; the first image quality index is adjusted using the first image block overlap degree as an adjustment coefficient.

[0081] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Figure 1 The image processing-based hydraulic equipment assembly and control method and specific examples in Embodiment 1 are also applicable to the image processing-based hydraulic equipment assembly and control system of this embodiment. Through the foregoing detailed description of the image processing-based hydraulic equipment assembly and control method, those skilled in the art can clearly understand the image processing-based hydraulic equipment assembly and control system of this embodiment; therefore, for the sake of brevity, it will not be described in detail here. As for the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and relevant details can be found in the method section.

[0082] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0083] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A hydraulic equipment assembly and control method based on image processing, characterized in that, The method includes: Acquire historical assembly vibration signals, which refer to the vibration signals of the hydraulic equipment during the first assembly stage in the historical assembly process; Multi-dimensional domain feature analysis was performed on the historical assembly vibration signals to obtain vibration signal domain feature information; The intelligent prediction model is invoked to predict the minimum vibration time point based on the vibration signal domain feature information, thereby obtaining the target time point; Obtain historical assembly lighting parameters, which refer to the lighting parameters of the first assembly stage of the hydraulic equipment during historical assembly processes; Extract the first historical ray parameter from the historical assembly ray parameters, and obtain the first historical acquisition image under the first historical ray parameter; An image quality evaluation function is introduced to obtain the first image quality index of the first historically acquired image, and when the first image quality index reaches the quality index limit, the first historical lighting parameters are used as preset lighting parameters. The lighting is set for the first assembly stage of the hydraulic equipment based on the preset lighting parameters, and the image acquisition device is started at the target time point. The image acquisition device is used to acquire assembly images of the first assembly stage to obtain a first assembly image, and the assembly of the first assembly stage is controlled based on the first assembly image. The method includes: Acquire training vibration signals and obtain the training vibration signal spectrum based on the principle of fast Fourier transform; The training vibration signal spectrum is analyzed to obtain the training spectral density, and the training target time point range corresponding to the peak value of the training signal in the training spectral density is back-matched. Embed the training target time point range into the first time point predictor; Obtain the training target time point of the training vibration signal, wherein the training target time point refers to the time point when the vibration is at its minimum in the training vibration signal; Supervised machine learning is performed on the training dataset constructed based on the vibration signal domain feature information and the training target time points to obtain a second time point predictor; The intelligent prediction model is constructed by combining the first time-point predictor and the second time-point predictor.

2. The hydraulic equipment assembly and control method based on image processing according to claim 1, characterized in that, The method includes: The training vibration signal is subjected to time-domain feature acquisition to obtain the time-domain features of the training signal; Frequency domain features of the training vibration signal are acquired by performing frequency domain feature acquisition; The time-domain features and frequency-domain features of the training signal constitute the vibration signal domain feature information.

3. The hydraulic equipment assembly and control method based on image processing according to claim 1, characterized in that, The expression for the image quality evaluation function is as follows: ； in, This refers to the first image quality index. This refers to the first image feature value of the first historically acquired image. This refers to the quality feedback coefficient. This refers to the coordinates in the first historically acquired image. DC coefficient of the cut image patch, This refers to the coordinates in the first historically acquired image. The AC coefficient of the segmented image patch, This refers to the influence factor of the DC coefficient on the feature value of the first image. This refers to the influence factor of the AC coefficient on the first image feature value, and .

4. The hydraulic equipment assembly and control method based on image processing according to claim 3, characterized in that, The method includes: The first historical image is cut to obtain a first cutting result, which includes multiple cut image blocks with coordinate labels; Extract the first cut image block from the plurality of cut image blocks with coordinate identifiers, wherein the first cut image block corresponds to the first coordinate; Obtain the first discrete cosine transform coefficient of the first cut image block, wherein the first discrete cosine transform coefficient includes a first DC coefficient and a first AC coefficient; Establish a first mapping relationship between the first coordinate and the first DC coefficient and the first AC coefficient; The image quality evaluation function calculates the first image quality index based on the first mapping relationship.

5. The hydraulic equipment assembly and control method based on image processing according to claim 3, characterized in that, The method further includes: The first saliency map of the first historical image is obtained based on the Itti-Koch visual saliency principle. Obtain the overlap degree between the target salient image block and the predetermined salient image block in the first salient image, wherein the target salient image block is a set of pixels in the first salient image whose salient values ​​satisfy a predetermined limit; The first image quality index is adjusted using the overlap of the first image block as an adjustment coefficient.

6. The hydraulic equipment assembly and control method based on image processing according to claim 1, characterized in that, Read the predetermined light characteristics and obtain the first historical light parameters based on the predetermined light characteristics, wherein the predetermined light characteristics include light source quality, light source intensity and light interference.

7. A hydraulic equipment assembly and control system based on image processing, characterized in that, The system comprises: steps for implementing the method according to any one of claims 1 to 6; A historical assembly vibration signal acquisition module is used to acquire historical assembly vibration signals, which refer to the vibration signals of the hydraulic equipment in the first assembly stage during the historical assembly process. A multi-dimensional domain feature analysis module is used to perform multi-dimensional domain feature analysis on the historical assembly vibration signal to obtain vibration signal domain feature information. The intelligent prediction model invocation module is used to invoke the intelligent prediction model to predict the minimum vibration time point of the vibration signal domain feature information and obtain the target time point. A historical assembly light parameter acquisition module is used to acquire historical assembly light parameters, which refer to the light parameters of the first assembly stage of the hydraulic equipment during historical assembly. A light parameter extraction module is used to extract the first historical light parameter from the historical assembly light parameters and obtain the first historical acquisition image under the first historical light parameter. The image quality assessment module is used to introduce an image quality assessment function to obtain a first image quality index of the first historically acquired image, and when the first image quality index reaches the quality index limit, the first historical lighting parameters are used as preset lighting parameters. A light setting module is used to set the light for the first assembly stage of the hydraulic equipment based on the preset light parameters, and to start the image acquisition device at the target time point. An image acquisition and assembly control module is used to acquire assembly images of the first assembly stage through the image acquisition device to obtain a first assembly image, and to control the assembly of the first assembly stage based on the first assembly image.

Images