Automatic control system and method for continuous production of real stone paint

Abstract

The invention relates to the field of intelligent control, and particularly discloses an automatic control system and method for continuous production of a real stone paint, and the method comprises the steps: collecting a color matching monitoring video and a reference color matching image in a color matching process of the real stone paint through a camera; and key frame extraction and image feature analysis are performed on the toning monitoring video and the reference toning image by adopting an image analysis and processing technology based on deep learning, so that a control result about whether to stop adding the colored sand or not is intelligently obtained. By means of the mode, the system can achieve automatic control over the color matching process of real stone paint production and accurately recognize the color difference and the change trend in the color matching process, so that the color consistency and accuracy are guaranteed, meanwhile, manual intervention is reduced, and the production efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent control, and more specifically, to an automatic control system and method for continuous production of real stone paint. Background Art

[0002] Real stone paint is a decorative paint with natural stone powder as the main raw material. It is made through a special process. It can create a decorative effect similar to marble or granite on the exterior walls of buildings. It has fireproof, waterproof and weather-resistant properties and is widely used in the decoration and protection of building exterior walls.

[0003] In the production process of real stone paint, color matching is a key step, the purpose of which is to ensure the consistency and accuracy of product color. Through precise color matching, customers' specific requirements for color can be met while ensuring the stability of color in mass production. Colored sand is a fine particle made of natural minerals or synthetic materials of different colors. They usually have a variety of colors and particle sizes. It can provide the texture and color of natural stone for real stone paint, making the coating appearance closer to real stone.

[0004] In the traditional color matching process of real stone paint production, colorists need to rely on personal experience and sensitivity to color to adjust the amount and type of colored sand added. However, since the color of real stone paint mostly depends on the color of natural colored sand, to achieve a specific color effect, it is necessary to accurately match the colored sand, which is more difficult using traditional methods.

[0005] Therefore, an optimized automatic control system for continuous production of real stone paint is expected. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an automatic control system and method for the continuous production of real stone paint, which collects the coloring monitoring video and reference coloring image in the coloring process of real stone paint through a camera, and uses image analysis and processing technology based on deep learning to perform key frame extraction and image feature analysis on the coloring monitoring video and reference coloring image, so as to intelligently obtain the control result of whether to stop adding colored sand. In this way, the system can realize the automatic control of the coloring process of real stone paint production, accurately identify the color differences and change trends in the coloring process, thereby ensuring the consistency and accuracy of the color, while reducing manual intervention and improving production efficiency.

[0007] According to one aspect of the present application, there is provided an automatic control system for continuous production of real stone paint, comprising:

[0008] A color grading monitoring video reference image acquisition module is used to acquire color grading monitoring video and reference color grading images during the color grading process of real stone paint through a camera;

[0009] A color adjustment state key frame extraction module, used for extracting key frames from the color adjustment monitoring video to obtain a time sequence of color adjustment state key frames;

[0010] A color feature capturing module for a color adjustment state is used to capture the color features of the time series of the color adjustment state key frames to obtain a time series of a color feature representation graph of the color adjustment state;

[0011] A feature enhancement fusion module is used to perform color feature enhancement on the time series of the toning state color feature representation graph to obtain the time series of the toning state color feature enhancement representation graph, and then perform pooling fusion on the time series of the toning state color feature enhancement representation graph to obtain the toning state color feature full time domain representation vector;

[0012] A reference toning color feature extraction module, used for capturing color features of the reference toning image to obtain a reference toning color feature representation vector;

[0013] A color feature classification representation module, used for fusing the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector to obtain a real stone paint color adjustment classification representation vector;

[0014] The control result generation module is used to obtain the control result based on the real stone paint color classification representation vector.

[0015] According to another aspect of the present application, there is provided an automatic control method for continuous production of real stone paint, comprising:

[0016] The color monitoring video and reference color images of the real stone paint coloring process are collected through the camera;

[0017] Extracting key frames from the color-adjusting monitoring video to obtain a time sequence of color-adjusting state key frames;

[0018] Capturing color features of the time series of the color adjustment state key frames to obtain a time series of color adjustment state color feature representation graphs;

[0019] After color feature enhancement is performed on the time series of the toning state color feature representation graph to obtain the time series of the toning state color feature enhanced representation graph, pooling fusion is performed on the time series of the toning state color feature enhanced representation graph to obtain the toning state color feature full time domain representation vector;

[0020] Capturing color features of the reference toned image to obtain a reference toned color feature representation vector;

[0021] The reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector are merged to obtain a real stone paint color adjustment classification representation vector;

[0022] Based on the real stone paint color classification representation vector, a control result is obtained.

[0023] Compared with the prior art, the present application provides an automatic control system and method for the continuous production of real stone paint, which uses a camera to collect the color monitoring video and reference color image during the color mixing process of the real stone paint, and uses image analysis and processing technology based on deep learning to perform key frame extraction and image feature analysis on the color monitoring video and reference color image, so as to intelligently obtain the control result of whether to stop adding colored sand. In this way, the system can realize the automatic control of the color mixing process of real stone paint production, accurately identify the color differences and change trends in the color mixing process, thereby ensuring the consistency and accuracy of the color, while reducing manual intervention and improving production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0025] Figure 1 It is a block diagram of an automatic control system for continuous production of real stone paint according to an embodiment of the present application.

[0026] Figure 2 This is a schematic diagram of the architecture of an automatic control system for continuous production of real stone paint according to an embodiment of the present application.

[0027] Figure 3 It is a block diagram of a feature enhancement fusion module in an automatic control system for continuous production of real stone paint according to an embodiment of the present application.

[0028] Figure 4 The present invention is a flowchart of an automatic control method for continuous production of real stone paint according to an embodiment of the present application. DETAILED DESCRIPTION

[0029] The embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, but rather these embodiments are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0030] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0031] It should be noted that the terms "first\second\third" involved in the embodiments of the present application are only used to distinguish similar objects, and do not represent a specific order for the objects. It is understandable that the specific order or sequence of "first\second\third" can be interchanged where permitted. It should be understood that the objects distinguished by "first\second\third" can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.

[0032] As shown in this disclosure, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular, but also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list, and the method or device may also include other steps or elements.

[0033] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0034] Real stone paint is a kind of architectural paint, which is mainly composed of natural stone powder, emulsion, pigment and various additives. It is a water-based environmentally friendly paint, named because it has the texture and decorative effect of natural stone. It has been widely used in the field of architectural decoration because of its environmental protection, beauty, durability and economy.

[0035] In the production process of real stone paint, color matching is a key step, the purpose of which is to ensure the consistency and accuracy of product color. Through precise color matching, customers' specific requirements for color can be met while ensuring the stability of color in mass production. Colored sand is a fine particle made of natural minerals or synthetic materials of different colors. They usually have a variety of colors and particle sizes. It can provide the texture and color of natural stone for real stone paint, making the coating appearance closer to real stone.

[0036] In the traditional color matching process of real stone paint production, colorists need to rely on personal experience and sensitivity to color to adjust the amount and type of colored sand added. However, since the color of real stone paint mostly depends on the color of natural colored sand, to achieve a specific color effect, it is necessary to accurately match the colored sand, which is more difficult using traditional methods.

[0037] Therefore, an optimized automatic control system for the continuous production of real stone paint is desired, which uses a camera to collect the color monitoring video and reference color image during the color mixing process of the real stone paint, and uses image analysis and processing technology based on deep learning to perform key frame extraction and image feature analysis on the color monitoring video and reference color image, so as to intelligently obtain the control result of whether to stop adding colored sand. In this way, the system can realize the automatic control of the color mixing process of real stone paint production, accurately identify the color differences and change trends in the color mixing process, thereby ensuring the consistency and accuracy of the color, while reducing manual intervention and improving production efficiency.

[0038] Figure 1 It is a block diagram of an automatic control system for continuous production of real stone paint according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of an automatic control system for continuous production of real stone paint according to an embodiment of the present application. Figure 1 and Figure 2As shown, according to the embodiment of the present application, the automatic control system 100 for continuous production of real stone paint includes: a color monitoring video reference image acquisition module 110, which is used to collect the color monitoring video and the reference color image in the coloring process of the real stone paint through a camera; a color state key frame extraction module 120, which is used to extract the key frames of the color monitoring video to obtain a time series of the color state key frames; a color state color feature capture module 130, which is used to capture the color features of the time series of the color state key frames to obtain the time series of the color feature representation diagram of the color state; a feature enhancement fusion module 140, which is used to perform color fusion on the time series of the color feature representation diagram of the color state After feature enhancement to obtain the time series of the toning state color feature enhancement representation graph, the time series of the toning state color feature enhancement representation graph is pooled and fused to obtain the toning state color feature full-time domain representation vector; a reference toning color feature extraction module 150 is used to capture the color features of the reference toning image to obtain a reference toning color feature representation vector; a color feature classification representation module 160 is used to fuse the reference toning color feature representation vector and the toning state color feature full-time domain representation vector to obtain a real stone paint color classification representation vector; and a control result generation module 170 is used to obtain a control result based on the real stone paint color classification representation vector.

[0039] In an embodiment of the present application, the coloring monitoring video reference image acquisition module 110 is used to collect the coloring monitoring video and reference coloring image in the coloring process of real stone paint through a camera. It should be understood that considering that the coloring monitoring video is a video stream of the real stone paint coloring process recorded in real time by a camera, the dynamic situation of color changes in the coloring process is captured. The reference coloring image is a predefined or known image with specific color attributes, representing the target color or color range, and is a reference standard for color accuracy in the coloring process. In the production of real stone paint, color consistency and accuracy are crucial. Based on this, in the technical solution of the present application, the coloring monitoring video and reference coloring image in the coloring process of real stone paint are collected by a camera, and the key frame extraction and image color feature analysis of the coloring monitoring video and the color feature analysis of the reference coloring image are performed, and the color can be monitored and adjusted in real time, so as to accurately control the amount of colored sand added, thereby helping to ensure the color quality of the final product and meet the standards of customers and the industry.

[0040] It is worth mentioning that the color grading monitoring video and the reference color grading image are usually collected using industrial or professional cameras, which need to have high resolution, high frame rate and good color reproduction capabilities to ensure color accuracy and consistency. In particular, in a specific embodiment of the present application, the acquisition can be performed by an industrial camera, which is designed for industrial applications and can provide stable and high-quality images, suitable for color-sensitive application scenarios. In addition, in order to capture the details of the real stone paint color grading process, high-definition cameras are necessary, which can provide sufficient image information for color analysis.

[0041] In an embodiment of the present application, the color adjustment state key frame extraction module 120 is used to extract key frames from the color adjustment monitoring video to obtain a time series of color adjustment state key frames. Accordingly, it is considered that the color adjustment monitoring video contains a large amount of video information and image frames in the color adjustment process. Key frames usually represent key moments or significant changes in the video, and extracting these frames helps to focus on analyzing important changes in the state during the video process. Therefore, in order to analyze and process the color adjustment state change information that has a key impact on color adjustment and is significant in a more fine-grained manner in the color adjustment monitoring video, in the technical solution of the present application, key frame extraction is performed on the color adjustment monitoring video to obtain a time series of color adjustment state key frames. That is, the time series of the color adjustment state key frames can better show the color change pattern and trend in the color adjustment process of real stone paint, and provide a basis and support for the subsequent extraction and analysis of color features.

[0042] In particular, in a specific embodiment of the present application, the time series of the color adjustment state key frames can be extracted in the following manner. First, by calculating the difference between adjacent frames of the color adjustment state, the frames in the color adjustment state video that have changed can be identified, which are usually key frames. Next, a threshold is set to determine when the difference between two frames is considered large enough to mark them as key frames. The above method can effectively extract key frames from the color adjustment monitoring video to obtain a time series of color adjustment state key frames.

[0043] In an embodiment of the present application, the color feature capture module 130 of the color-adjusting state is used to capture the color features of the time series of the color-adjusting state key frames to obtain the time series of the color feature representation diagram of the color-adjusting state. Specifically, in an embodiment of the present application, the color feature capture module of the color-adjusting state is used to: input the time series of the color-adjusting state key frames into the color feature capture model based on the convolutional neural network to obtain the time series of the color feature representation diagram of the color-adjusting state. Accordingly, considering that each color-adjusting state key frame in the time series of the color-adjusting state key frames contains the color change and feature information of the real stone paint color-adjusting process at different time points, and the convolutional neural network is good at automatically learning features from images, and can capture the key features of color changes, and has a strong perception ability for color features. Based on this, in the technical solution of the present application, the time series of the color-adjusting state key frames is input into the color feature capture model based on the convolutional neural network to obtain the time series of the color feature representation diagram of the color-adjusting state, so that the complex feature representation contained in the time series of the color-adjusting state key frames can be learned, which helps to understand the color changes of the real stone paint body color process at each time point in more detail.

[0044] In an embodiment of the present application, the feature enhancement and fusion module 140 is used to perform color feature enhancement on the time series of the toning state color feature representation graph to obtain the time series of the toning state color feature enhanced representation graph, and then perform pooling and fusion on the time series of the toning state color feature enhanced representation graph to obtain the full time domain representation vector of the toning state color feature. Figure 3 : is a block diagram of a feature enhancement fusion module in an automatic control system for continuous production of real stone paint according to an embodiment of the present application. Specifically, in the embodiment of the present application, Figure 3 As shown, the feature enhancement and fusion module 140 includes: a color feature enhancement unit 141, which is used to input the time series of the toning state color feature representation graph into the color feature enhancement module based on the bidirectional attention mechanism to obtain the time series of the toning state color feature enhancement representation graph; a color feature global pooling processing unit 142, which is used to perform global pooling processing of each feature matrix along the channel dimension on each toning state color feature enhancement representation graph in the time series of the toning state color feature enhancement representation graph to obtain the time series of the toning state color feature enhancement representation vector; a full-time domain feature fusion unit 143, which is used to perform full-time domain feature fusion on the time series of the toning state color feature enhancement representation vector to obtain the full-time domain representation vector of the toning state color feature.

[0045] Specifically, the color feature enhancement unit 141 is used to input the time series of the toning state color feature representation diagram into the color feature enhancement module based on the bidirectional attention mechanism to obtain the time series of the toning state color feature enhancement representation diagram. It should be understood that, considering that each toning state color feature representation diagram in the time series of the toning state color feature representation diagram has different toning state color features in different channels, it contains important key color feature information and also contains background information and noise that are not related to color feature processing. Based on this, in order to focus more on the feature information that has an effect on the color feature change in the toning process, in the technical solution of the present application, the time series of the toning state color feature representation diagram is input into the color feature enhancement module based on the bidirectional attention mechanism to obtain the time series of the toning state color feature enhancement representation diagram. It is worth mentioning that the attention mechanism can highlight the most important areas or features in the image and enhance the model's recognition ability of key color features, while the bidirectional attention mechanism can simultaneously consider the context feature information of the color features of the toning state color feature representation diagram in different directions to provide a more comprehensive color representation.

[0046] Specifically, the color feature global pooling processing unit 142 is used to perform global pooling processing of each feature matrix along the channel dimension on each color-adjusting state color feature enhancement representation graph in the time series of the color-adjusting state color feature enhancement representation graph to obtain a time series of color-adjusting state color feature enhancement representation vectors. Accordingly, each color-adjusting state color feature enhancement representation graph in the time series of the color-adjusting state color feature enhancement representation graph contains important feature information about the color state. Therefore, in order to capture and mine the key features of the color at each time point while reducing redundant information, in the technical solution of the present application, each color-adjusting state color feature enhancement representation graph in the time series of the color-adjusting state color feature enhancement representation graph is subjected to global pooling processing of each feature matrix along the channel dimension to obtain a time series of color-adjusting state color feature enhancement representation vectors.

[0047] Specifically, the full-time-domain feature fusion unit 143 is used to perform full-time-domain feature fusion on the time series of the toning state color feature enhancement representation vector to obtain the full-time-domain representation vector of the toning state color feature. Specifically, in an embodiment of the present application, the full-time-domain feature fusion unit is used to: arrange the time series of the toning state color feature enhancement representation vector in time series to obtain the full-time-domain representation vector of the toning state color feature. It should be understood that, considering that each toning state color feature enhancement representation vector in the time series of the toning state color feature enhancement representation vector represents the color feature information of the toning state at a time point, therefore, in order to more comprehensively and comprehensively integrate the overall change of the toning in the entire time domain, in the technical solution of the present application, the time series of the toning state color feature enhancement representation vector is arranged in time series to obtain the full-time-domain representation vector of the toning state color feature, so that the trend and pattern of the color change of the real stone paint within the entire time range can be captured by considering the characteristics of the entire time series.

[0048] In an embodiment of the present application, the reference coloring color feature extraction module 150 is used to capture the color features of the reference coloring image to obtain a reference coloring color feature representation vector. Specifically, in an embodiment of the present application, the reference coloring color feature extraction module is used to: input the reference coloring image into the color feature capture model based on the convolutional neural network to obtain the reference coloring color feature representation vector. Similarly, considering that the reference coloring image contains the color implicit feature information in the reference coloring image, and the convolutional neural network (CNN) is particularly good at automatically learning complex implicit feature representations from images, which is crucial for understanding the subtle differences in color. Based on this, in the technical solution of the present application, the reference coloring image is input into the color feature capture model based on the convolutional neural network to obtain the reference coloring color feature representation vector, so that the color features implied in the reference coloring image can be extracted and captured for comparison and matching with the image in the coloring process, thereby improving the accuracy of the true stone paint coloring process.

[0049] In an embodiment of the present application, the color feature classification representation module 160 is used to fuse the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector to obtain the real stone paint color adjustment classification representation vector. Specifically, in an embodiment of the present application, the color feature classification representation module is used to: perform position differential calculation on the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector to obtain the real stone paint color adjustment classification representation vector. It should be understood that, considering that the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector respectively express the color feature information of the reference color adjustment and the color feature information of the color adjustment process within the full time range. Therefore, in order to more accurately quantify the difference between the color features of the two and provide a basis for colored sand control, in the technical solution of the present application, the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector are calculated by position differential calculation to obtain the real stone paint color adjustment classification representation vector. That is, by performing positional differentiation between the two, the reference image and the color adjustment state are aligned in space, making the color comparison more accurate, thereby facilitating accurate color control.

[0050] In the embodiment of the present application, the control result generation module 170 is used to obtain the control result based on the real stone paint color classification representation vector. Specifically, in the embodiment of the present application, the control result generation module includes: a perspective mapping matching unit, which is used to perform perspective mapping matching fusion based on feature granularity association on the reference color feature representation vector and the color state color feature full time domain representation vector to obtain the real stone paint color classification representation vector; a control unit, which is used to input the real stone paint color classification representation vector into the color sand controller based on the classifier to obtain the control result of whether to stop adding color sand.

[0051] Specifically, the perspective mapping matching unit is used to perform perspective mapping matching fusion based on feature granularity association on the reference color toning color feature representation vector and the color toning state color feature full time domain representation vector to obtain the real stone paint color classification representation vector. In particular, in the technical solution of the present application, in the technical solution of the present application, it is considered that the reference color toning color feature representation vector and the color toning state color feature full time domain representation vector are extracted from the static reference color toning image and the dynamic color toning monitoring video respectively. Due to the significant differences in their sources, properties and extraction methods, feature heterogeneity is caused. Specifically, the reference color toning color feature representation vector and the color toning state color feature full time domain representation vector may exist on different manifolds, which means that they have different distribution laws in their respective spaces. Directly splicing them together may result in the final feature vector not having a clear manifold in the new space, that is, the feature manifold smoothness is insufficient. Since the two feature vectors come from different sources, they may not be well mapped to each other in the fused feature space, that is, one type of feature may not accurately reflect the feature changes of another type, resulting in increased uncertainty in feature expression. In order to solve this problem, in the technical solution of the present application, the reference toning color feature representation vector and the toning state color feature full-time domain representation vector are subjected to viewing angle mapping matching fusion based on feature granularity association.

[0052] More specifically, the perspective mapping matching unit is used to: construct a real stone paint color adjustment distance topological matrix between the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector; construct a real stone paint color adjustment feature value granularity association matrix between the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector; perform topological association modulation on the real stone paint color adjustment feature value granularity association matrix based on the real stone paint color adjustment distance topological matrix to obtain a real stone paint color adjustment topological modulation association matrix; use the real stone paint color adjustment topological modulation association matrix as the perspective co-projection space, project the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector to the perspective co-projection space to obtain a perspective modulation reference color adjustment color feature representation vector and a perspective modulation color adjustment state color feature full time domain representation vector; fuse the perspective modulation reference color adjustment color feature representation vector and the perspective modulation color adjustment state color feature full time domain representation vector to obtain the real stone paint color adjustment classification representation vector.

[0053] Specifically, the category harmony matching unit is used to: perform perspective mapping matching fusion based on feature granularity association on the reference color feature representation vector and the color feature full-time domain representation vector of the color state according to the following formula to obtain the real stone paint color classification representation vector; wherein the formula is:

[0054] M t =D(V1 ,V 2 )

[0055]

[0056] V c =α·V 1 ′+β·V 2 '

[0057] Among them, V 1 represents the reference color feature representation vector, V 2 The color feature of the color adjustment state is represented by the full time domain vector, D(V 1 ,V 2 ) is the real stone paint toning distance topological matrix between the reference toning color feature representation vector and the toning state color feature full time domain representation vector, that is, d i,j =d(v 1i ,v 2j ), d i,j is the eigenvalue of the position (i, j) in the real stone paint toning distance topological matrix, d(v 1i ,v 2j ) represents the distance between the characteristic value of the i-th position of the reference color adjustment color feature representation vector and the characteristic value of the j-th position of the color adjustment state color feature full time domain representation vector, and V 1 and V 2 are all column vectors, M t represents the real stone paint toning distance topological matrix, T represents the transpose of the vector, Represents matrix multiplication, M v represents the granularity correlation matrix of the color eigenvalue of real stone paint, M c represents the topological modulation correlation matrix of real stone paint color matching, V 1 ′ represents the reference color feature representation vector of the viewing angle modulation, V 2 ′ represents the full-time domain representation vector of the color feature of the viewing angle modulation color adjustment state, α and β represent weighted hyperparameters, V c Represents the real stone paint color classification representation vector.

[0058] Based on this, in the technical solution of the present application, the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector are subjected to perspective mapping matching fusion based on feature granularity association, which first constructs a real stone paint color adjustment distance topological matrix between the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector. This step utilizes the idea of ​​graph theory, taking the eigenvalues ​​of each position in the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector as nodes, and using the distance metric function to determine the low-dimensional embedding expression of the edges between nodes, so as to construct a feature expression that can quantify the topological association relationship between the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector. This step is crucial to understanding the spatial distribution of data points and their mutual connections, and also provides basic structural information for subsequent steps.

[0059] Next, a real stone paint color tracing feature value granularity association matrix is ​​constructed between the reference color tracing color feature representation vector and the color tracing state color feature full-time domain representation vector. This step focuses on exploring the correlation between the internal attributes of the feature vector. In a specific example, the product between the transposed vectors of the reference color tracing color feature representation vector and the color tracing state color feature full-time domain representation vector is calculated to obtain the real stone paint color tracing feature value granularity association matrix. Furthermore, the real stone paint color tracing feature value granularity association matrix is ​​topologically associated modulated based on the real stone paint color tracing distance topological matrix to obtain the real stone paint color tracing topological modulation association matrix. That is, the feature value granularity association information of the reference color tracing color feature representation vector and the color tracing state color feature full-time domain representation vector is driven to wander on the distance topological map to construct a low-dimensional modulation perspective co-projection space domain for mapping the reference color tracing color feature representation vector and the color tracing state color feature full-time domain representation vector to a continuous high-dimensional regression space attribute.

[0060] Next, the real stone paint color topology modulation association matrix is ​​used as the perspective co-projection space, and the reference color feature representation vector and the color state color feature full time domain representation vector are projected into the perspective co-projection space to obtain the perspective modulation reference color feature representation vector and the perspective modulation color state color feature full time domain representation vector. That is, the key to this technical solution is to find an implicit third space with modulation capability, in which as much information of the original high-dimensional data as possible is retained, so that the converted new feature vector not only reduces the dimension, but also maintains the relative position relationship between the original features.

[0061] Finally, the perspective modulation reference color feature representation vector and the perspective modulation color state color feature full-time domain representation vector are fused to obtain the real stone paint color classification representation vector. Fusion can be achieved in a variety of forms, ranging from simple arithmetic averages to complex ensemble learning algorithms, depending on the application scenario and the goals pursued. The final real stone paint color classification representation vector combines the main features of the two sets of input data and provides a more comprehensive data representation, which is not only conducive to improving the performance of machine learning models, but also promotes the ability of cross-modal or multi-source data analysis, thereby significantly enhancing the understanding and utilization efficiency of complex data structures.

[0062] Specifically, the control unit is used to input the real stone paint color classification representation vector into the colored sand controller based on the classifier to obtain the control result of whether to stop adding colored sand. That is, the real stone paint color classification representation vector is classified and processed to intelligently obtain the control result of whether to stop adding colored sand. In this way, the system can realize the automatic control of the real stone paint coloring process, accurately identify color differences and change trends, ensure color consistency and accuracy, reduce manual intervention, and thus improve production efficiency.

[0063] It is worth mentioning that the classifier-based colored sand controller can automatically determine whether it is necessary to stop adding colored sand according to the real stone paint color classification representation vector, thereby reducing the need for manual monitoring and decision-making. That is, by accurately analyzing the color difference, the classifier-based colored sand controller can determine when the required color matching degree is achieved, thereby accurately controlling the amount of colored sand added. In this way, production efficiency can be improved and delays caused by waiting for manual decisions can be reduced.

[0064] In summary, the automatic control system 100 for continuous production of real stone paint based on the embodiment of the present application is explained, which uses a camera to collect the color monitoring video and reference color image during the coloring process of the real stone paint, and uses image analysis and processing technology based on deep learning to perform key frame extraction and image feature analysis on the color monitoring video and reference color image, so as to intelligently obtain the control result of whether to stop adding colored sand. In this way, the system can realize the automatic control of the coloring process of real stone paint production, accurately identify the color differences and change trends in the coloring process, thereby ensuring the consistency and accuracy of the color, while reducing manual intervention and improving production efficiency.

[0065] Figure 4 Flow chart of the automatic control method for continuous production of real stone paint according to an embodiment of the present application. Figure 4As shown, according to the embodiment of the present application, the automatic control method for continuous production of real stone paint includes: S110, collecting the coloring monitoring video and the reference coloring image in the coloring process of the real stone paint through a camera; S120, extracting the key frames of the coloring monitoring video to obtain the time series of the key frames of the coloring state; S130, capturing the color features of the time series of the key frames of the coloring state to obtain the time series of the color feature representation diagram of the coloring state; S140, after performing color feature enhancement on the time series of the color feature representation diagram of the coloring state to obtain the time series of the color feature enhancement representation diagram of the coloring state, pooling and fusion are performed on the time series of the color feature enhancement representation diagram of the coloring state to obtain the color feature full-time domain representation vector of the coloring state; S150, capturing the color features of the reference coloring image to obtain the reference color feature representation vector; S160, fusing the reference color feature representation vector and the color feature full-time domain representation vector of the coloring state to obtain the real stone paint color classification representation vector; S170, obtaining the control result based on the real stone paint color classification representation vector.

[0066] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned automatic control method for continuous production of real stone paint have been described in detail above. Figures 1 to 3 The automatic control system for continuous production of true stone paint has been described in detail, and therefore, its repeated description will be omitted.

[0067] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital video disc (DVD)), or a semiconductor medium (eg, a solid state disk (SSD)).

[0068] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0069] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0070] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0071] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit of the technical solution of the present invention.

Claims

1. An automatic control system for continuous production of real stone paint, characterized in that: include: A color grading monitoring video reference image acquisition module is used to acquire color grading monitoring video and reference color grading images during the color grading process of real stone paint through a camera; A color adjustment state key frame extraction module, used for extracting key frames from the color adjustment monitoring video to obtain a time sequence of color adjustment state key frames; A color feature capturing module for a color adjustment state is used to capture the color features of the time series of the color adjustment state key frames to obtain a time series of a color feature representation graph of the color adjustment state; A feature enhancement fusion module is used to perform color feature enhancement on the time series of the toning state color feature representation graph to obtain the time series of the toning state color feature enhancement representation graph, and then perform pooling fusion on the time series of the toning state color feature enhancement representation graph to obtain the toning state color feature full time domain representation vector; A reference toning color feature extraction module, used for capturing color features of the reference toning image to obtain a reference toning color feature representation vector; A color feature classification representation module, used for fusing the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector to obtain a real stone paint color adjustment classification representation vector; The control result generation module is used to obtain the control result based on the real stone paint color classification representation vector.

2. The automatic control system for continuous production of real stone paint according to claim 1, characterized in that: The color feature capturing module of the color adjustment state is used for: The time series of the color adjustment state key frames is input into a color feature capture model based on a convolutional neural network to obtain the time series of the color feature representation graph of the color adjustment state.

3. The automatic control system for continuous production of real stone paint according to claim 2 is characterized in that: The feature enhancement fusion module, include: A color feature enhancement unit, configured to input the time series of the color feature representation graph of the toning state into a color feature enhancement module based on a bidirectional attention mechanism to obtain the time series of the color feature enhancement representation graph of the toning state; A color feature global pooling processing unit is used to perform global pooling processing on each feature matrix of each color-adjusting state color feature enhancement representation graph in the time series of the color-adjusting state color feature enhancement representation graph along the channel dimension to obtain a time series of the color-adjusting state color feature enhancement representation vector; The full-time domain feature fusion unit is used to perform full-time domain feature fusion on the time series of the toning state color feature enhancement representation vector to obtain the full-time domain representation vector of the toning state color feature.

4. The automatic control system for continuous production of real stone paint according to claim 3 is characterized in that: The full time domain feature fusion unit is used to: The time series of the enhanced representation vectors of the toning state color features are arranged in time sequence to obtain the full-time domain representation vector of the toning state color features.

5. The automatic control system for continuous production of real stone paint according to claim 4, characterized in that: The reference color feature extraction module is used to: The reference color-adjusted image is input into the convolutional neural network-based color feature capture model to obtain the reference color-adjusted color feature representation vector.

6. The automatic control system for continuous production of real stone paint according to claim 5, characterized in that: The color feature classification and representation module is used to: The reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector are calculated by position difference to obtain the real stone paint color adjustment classification representation vector.

7. The automatic control system for continuous production of real stone paint according to claim 6, characterized in that: The control result generating module comprises: A viewing angle mapping matching unit is used to perform viewing angle mapping matching fusion based on feature granularity association on the reference toning color feature representation vector and the toning state color feature full-time domain representation vector to obtain a true stone paint toning classification representation vector; The control unit is used to input the real stone paint color classification representation vector into the color sand controller based on the classifier to obtain the control result of whether to stop adding the color sand.

8. The automatic control system for continuous production of real stone paint according to claim 7, characterized in that: The perspective mapping matching unit is used to: Constructing a real stone paint color adjustment distance topological matrix between the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector; Constructing a real stone paint color adjustment feature value granularity association matrix between the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector; Based on the real stone paint coloring distance topological matrix, the real stone paint coloring characteristic value granularity association matrix is ​​topologically modulated to obtain a real stone paint coloring topological modulation association matrix; Taking the real stone paint color adjustment topology modulation association matrix as the perspective co-projection space, projecting the reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector into the perspective co-projection space to obtain the perspective modulation reference color adjustment color feature representation vector and the perspective modulation color adjustment state color feature full time domain representation vector; The viewing angle modulation reference color feature representation vector and the viewing angle modulation color feature full-time domain representation vector are fused to obtain the real stone paint color classification representation vector.

9. An automatic control method for continuous production of real stone paint, characterized in that: include: The color monitoring video and reference color images of the real stone paint coloring process are collected through the camera; Extracting key frames from the color-adjusting monitoring video to obtain a time sequence of color-adjusting state key frames; Capturing color features of the time series of the color adjustment state key frames to obtain a time series of color adjustment state color feature representation graphs; After color feature enhancement is performed on the time series of the toning state color feature representation graph to obtain the time series of the toning state color feature enhanced representation graph, pooling fusion is performed on the time series of the toning state color feature enhanced representation graph to obtain the toning state color feature full time domain representation vector; Capturing color features of the reference toned image to obtain a reference toned color feature representation vector; The reference color adjustment color feature representation vector and the color adjustment state color feature full time domain representation vector are merged to obtain a real stone paint color adjustment classification representation vector; Based on the real stone paint color classification representation vector, a control result is obtained.

10. The automatic control method for continuous production of real stone paint according to claim 9, characterized in that: Capturing the color features of the reference toning image to obtain a reference toning color feature representation vector includes: The reference color-adjusted image is input into the convolutional neural network-based color feature capture model to obtain the reference color-adjusted color feature representation vector.

Images