A robust image restoration method for surface quality inspection of automotive parts

Abstract

The present invention provides a robust image restoration method for surface quality inspection of automotive parts. By collecting point cloud data of the surface of automotive parts and constructing a high-dimensional structured representation, the method combines a noise separation data model and multidimensional data decomposition technology to achieve accurate separation of target data and noise. This method constructs an optimization model containing a fidelity term, a regularization term, and a noise reduction term, optimizes the error between the restored data and the collected data, and performs data initialization and feature enhancement through a feature extraction module, thereby improving the accuracy and geometric consistency of the restored data. The model is solved using an iterative optimization method to ultimately obtain noise-free restored point cloud data. The present invention can effectively suppress noise in complex industrial environments, improve restoration accuracy and robustness, and provide efficient and reliable technical support for surface quality inspection of automotive parts.

Description

Technical Field

[0001] The present invention relates to the field of industrial vision, and in particular to a robust image restoration method for surface quality inspection of automotive parts. The method aims to improve the accuracy and applicability of surface inspection of automotive parts in complex industrial environments through point cloud data modeling and restoration technology. Background Art

[0002] As the pace of global industrialization accelerates, manufacturing technology continues to innovate and upgrade, especially in the automotive manufacturing sector. Product quality, production efficiency, and automation levels have become important indicators for measuring core competitiveness. The surface quality of automotive parts, as an important measure of the quality and performance of the entire vehicle, plays a key role in improving the appearance consistency and product life of the vehicle. Therefore, higher requirements are placed on the quality inspection of the surface of automotive parts. Traditional manual inspection methods are difficult to meet the needs of modern industrialized production in terms of efficiency and accuracy. Especially in large-scale, high-speed production lines, manual inspection is often limited by the subjectivity of manual judgment and the complexity of the inspection environment, resulting in problems such as missed inspections, false detections, and low efficiency. In recent years, the emergence of industrial visual inspection technology has provided a new solution for surface quality inspection of automotive parts, and has gradually become an important means to replace manual inspection.

[0003] Industrial visual inspection systems typically use 3D point cloud data to obtain the geometric shape and feature information of component surfaces. Point cloud data is a data format that can describe the geometric structure of objects with high precision and is widely used in fields such as autonomous driving, reverse engineering, virtual reality, and industrial inspection. By analyzing and processing 3D point cloud data on the surfaces of automotive parts, efficient identification of surface defects such as surface flaws, cracks, and dents can be achieved. However, point cloud data is often interfered with by a variety of complex environmental factors during the acquisition process, such as noise, uneven lighting, occlusion, and reflection. This leads to a decrease in the quality of the collected data, generates large noise or distortion, and directly affects the subsequent inspection accuracy and effectiveness. Especially in actual industrial environments, factors such as changes in ambient lighting, mechanical vibration, and sensor accuracy can have a negative impact on the quality of point cloud data acquisition, making it impossible for the inspection system to effectively cope with the diverse noise interference on the surfaces of complex components, reducing inspection accuracy and consistency.

[0004] Point cloud data restoration methods have been increasingly applied to automotive component surface quality inspection. Traditional point cloud restoration methods typically use global or local smoothing techniques to eliminate noise and improve data quality. However, these methods often struggle to strike a good balance between noise suppression and detail preservation in scenes with complex geometric structures and textures. While existing global smoothing techniques can effectively eliminate large-scale noise, they can blur edge details or even misidentify them as noise and remove them, resulting in the loss of key geometric features. Local smoothing methods, while able to preserve detail to a certain extent, often perform poorly when dealing with large areas of unstructured noise, making them ineffective in meeting the high-precision and high-stability requirements of automotive component surface quality inspection. Furthermore, complex automotive component surfaces often exhibit high reflectivity, curved features, and a rich variety of subtle structures and textures. These factors further complicate point cloud data restoration and place higher demands on the applicability and robustness of data restoration technologies.

[0005] In order to strike a balance between noise suppression and detail preservation, some point cloud restoration algorithms have been improved based on different mathematical models and data features. Common methods include using sparse representation, low-rank matrix decomposition and deep learning technology to restore point cloud data. Among them, sparse representation and low-rank decomposition methods are based on the sparsity and low-rank characteristics of noise and signals. By constructing specific mathematical models, the noise components in the data are separated from the real signal, thereby removing noise and improving the quality of point cloud data. However, these methods often require specific assumptions about the type of noise, data characteristics, etc., such as assuming that the noise has a sparse distribution or presents a Gaussian distribution. This results in limited applicability when dealing with diverse noises in complex industrial scenarios and inability to flexibly respond to different types of noise. The low-rank matrix decomposition method has high computational complexity when processing large-scale data, and it is difficult to meet the requirements of real-time detection.

[0006] In recent years, deep learning has achieved remarkable results in image restoration and noise suppression. Several deep neural network-based image restoration methods have been applied to point cloud data restoration. Deep learning methods typically construct deep models, such as convolutional neural networks or generative adversarial networks, and utilize large amounts of labeled data for training, enabling the models to achieve a good balance between denoising and detail preservation. However, deep learning methods also face numerous challenges when applied to real-world industrial scenarios. First, deep learning models typically rely on large amounts of labeled data, which is particularly difficult to obtain in point cloud data. Second, deep learning models often lack the ability to accurately represent details when processing the complex surfaces of automotive parts, and are prone to missing key geometric features. Furthermore, the training and inference processes of deep learning models are computationally intensive, making them difficult to meet the requirements of real-time detection.

[0007] Existing point cloud restoration methods still have many deficiencies in their applicability and robustness in complex scenarios. In particular, in the surface inspection scenario of automotive parts, the collected point cloud data has complex surface geometry and high reflectivity, which places higher demands on the applicability of the restoration method. Traditional point cloud restoration methods often struggle to simultaneously achieve effective noise suppression and detail preservation in such complex situations, which can easily lead to poor restoration results. Specifically, the deficiencies of existing methods are mainly reflected in the following aspects:

[0008] 1. Poor Applicability: Existing methods often rely on specific noise assumptions or geometric models, assuming a specific noise distribution or processing only data with regular geometric shapes. However, for automotive parts with complex shapes, curved surfaces, and multiple levels of detail, such assumptions are very limited and difficult to adapt to diverse industrial environments.

[0009] 2. Loss of detail: Existing point cloud restoration methods mostly use global or local smoothing techniques to remove noise. While this approach can improve data quality to a certain extent, it can lead to blurred edge details or misidentification and removal of details as noise. This loss of detail can severely impact subsequent inspection accuracy, especially on surfaces with complex geometric shapes like automotive parts.

[0010] 3. High computational complexity: As point cloud data volumes increase, traditional restoration algorithms often struggle to meet real-time requirements. This is particularly true in industrial production, where large-scale, real-time inspections are required. Existing methods struggle to achieve high-quality restoration while maintaining efficiency.

[0011] In summary, in the field of surface quality inspection of automotive parts, existing point cloud data restoration methods currently have certain limitations in terms of applicability, detail preservation capabilities, and computational efficiency. In particular, when faced with complex industrial environments and diverse noise interference, traditional methods are unable to meet the needs of industrial production for high-precision and high-consistency inspection. To this end, a more robust, adaptable, and efficient point cloud data restoration technology is urgently needed to improve the accuracy and applicability of surface quality inspection of automotive parts and provide reliable data support for subsequent inspection and quality control. This restoration technology not only needs to have the ability to adaptively suppress noise in complex noisy environments, but also needs to be able to effectively preserve the geometric features and detailed information of the surface of automotive parts, thereby ensuring that the quality of the restored data meets the requirements of high-precision inspection. Summary of the Invention

[0012] In order to solve the above problems in the prior art, the present invention proposes a robust image restoration method for surface quality inspection of automobile parts, which is characterized by comprising the following steps:

[0013] Collecting surface point cloud data of automobile parts containing three-dimensional geometric information, and constructing the surface point cloud data of automobile parts into a high-dimensional structured representation;

[0014] The point cloud data includes target data and different types of noise components, and a data model for noise separation is established based on the point cloud data;

[0015] The point cloud data is processed using a multidimensional data decomposition method, and the signal and noise are separated through a noise separation data model.

[0016] Construct an optimization model that includes fidelity terms, regularization terms, and noise reduction terms to optimize the error between the restored data and the collected data;

[0017] The data is initialized and enhanced through a feature extraction module; the optimization model is solved using an iterative optimization method to ultimately obtain noise-free restored point cloud data.

[0018] Using a third-order tensor Model the point cloud data of the surface of automobile parts and define the collected noise pollution point cloud data as tensor Y, satisfying Y=X+N g +S, where N g represents Gaussian noise, S represents sparse noise, and X represents the target tensor of noise-free point cloud data in the ideal case.

[0019] The point cloud data is represented as a high-dimensional target tensor X = Φ(G), where G = {G (1) ,G (2) ,…,G (n)} is a third-order core tensor sequence, and each core tensor The third-order tensor ring decomposition structure is used to adaptively adjust the weights of different dimensions to achieve robust processing of different noise types and data missing situations, where r n-1 and r n represents the rank of the tensor ring; I n is the number of features in the nth dimension.

[0020] The point cloud image is restored by optimizing the model, and the optimization model is:

[0021]

[0022] in, is the fidelity term, g(X) is the non-local denoising function; the parameter Y represents the collected point cloud data, which is the original data containing noise and serves as the input of the model; the parameter X is the restored point cloud data; the sparse noise S represents the non-structural noise in the collected data; the tensor decomposition result Φ(G) is the structured expression of the point cloud data; the parameter β1 controls the degree of matching between the restored data X and the tensor decomposition result Φ(G), the parameter β2 is used to adjust the regularization strength of the sparse noise S, and the parameter μ affects the balance between the non-local denoising function g(X) in denoising and detail preservation.

[0023] Using deep convolutional neural network f θ (X0) performs feature extraction and data initialization, where the input is a randomly initialized tensor X0, and the image edge and texture detail information is extracted through convolution operation, and the regional threshold is calculated using grayscale average. Complete image reconstruction; where u1 and u2 represent the grayscale averages of adjacent areas respectively.

[0024] The optimization model is solved step by step using the alternating direction method of multipliers (ADMM), including the following steps:

[0025] Solve for G (1) Sub-problem: The optimization objective is in and is the modal expansion of the core tensor;

[0026] Solve for G (2) Sub-problem: Update

[0027] Solve for G (3) Sub-problem: Update

[0028] Solve the subproblem of S: the optimization goal is And update S (t+1) =shr ink(Y-Φ(G) t ,μ);

[0029] Solve the subproblem of X: Update

[0030] Solve the subproblem of θ: Update

[0031] Among them: G (1) ,G (2) ,G (3) It is a third-order core tensor; it expresses the local features and geometric structure of point cloud data through tensor ring decomposition;

[0032] Tensor G(1) The second mode expansion and G (2) , G (3) The multimodal expansion results of

[0033] fo ld2: modal contraction operation;

[0034] The expanded form of the tensor X in different modes; representing the matrices after the first mode, second mode, and third mode of X are expanded respectively;

[0035] t+1 represents the iterative update;

[0036] G (t+1) : tensor decomposition result updated in the t+1th iteration;

[0037] S (t+1) : sparse noise after optimization in the t+1th iteration;

[0038] X (t+1) : The restored tensor calculated in the t+1th iteration;

[0039] Y: collected point cloud data;

[0040] Φ(G) t : tensor decomposition result of the current iteration;

[0041] The square of the Froben i us norm;

[0042] S: sparse noise;

[0043] shr ink(x,μ): soft threshold function;

[0044] μ: threshold parameter of soft threshold function;

[0045] β1,β2: penalty parameters of the optimization model;

[0046] g(X): non-local noise reduction function;

[0047] f θ (X0): Feature tensor extracted by the deep network;

[0048] θ: weight parameter of deep network;

[0049] X0: Initial input tensor.

[0050] Beneficial effects:

[0051] The present invention provides a robust image restoration method for surface quality inspection of automotive parts. By constructing point cloud data into a high-dimensional structured representation, combined with a noise separation data model and multidimensional data decomposition technology, the method achieves accurate separation of target data from different types of noise. By introducing an optimization model of fidelity terms, regularization terms, and noise reduction terms, the present invention significantly improves the accuracy and geometric consistency of the restored data, while effectively suppressing noise interference. A feature extraction module is used for data initialization and feature enhancement to enrich the detailed expression of the restored data. Through an iterative optimization method, the convergence and efficiency of the restoration process are further ensured. The present invention can perform high-quality restoration of point cloud data on the surface of automotive parts in complex industrial environments, significantly improve detection accuracy and robustness, and provide strong technical support for industrial quality inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application, but do not constitute an improper limitation of the present invention. In the drawings:

[0053] Figure 1 A flow chart of the method of the present invention is shown. DETAILED DESCRIPTION

[0054] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The exemplary embodiments and descriptions are only used to explain the present invention but are not intended to limit the present invention.

[0055] This embodiment provides a method for restoring point cloud data of noise-contaminated automotive component surfaces through tensor modeling and decomposition. This method uses third-order tensors to perform detailed modeling and restoration of point cloud data, resolving the existing technical challenge of balancing noise suppression and detail preservation in complex industrial environments. The following detailed description of the method's implementation process is provided with reference to the accompanying figures.

[0056] Step 1: Tensor modeling of automobile parts surface point cloud images

[0057] The present invention adopts the third-order tensor To represent point cloud data of automotive component surfaces, M, N, and P represent the spatial length, width, and number of channels of the tensor, respectively. This constructs a high-dimensional data structure with spatial and channel dimensions. Through tensor modeling, the geometric information and texture features of automotive component surfaces can be effectively expressed in high-dimensional space, providing a precise structural description for subsequent data restoration.

[0058] Specifically, the collected point cloud data is defined as a tensor Y. During the actual collection process, the tensor is often contaminated by various noises in the industrial environment, resulting in data distortion. In order to accurately characterize such noise, the present invention assumes that the point cloud image data Y can be decomposed into three parts: the target tensor X, the Gaussian noise N g and sparse noise S, that is, satisfying the relationship Y = X + N g +S. Among them, the target tensor X represents the ideal noise-free point cloud data, which is used to truly reflect the structural information of the surface of automobile parts; Gaussian noise N g This noise component, introduced during the acquisition process by random factors such as ambient lighting and sensor jitter, manifests as global interference. Sparse noise, S, represents strong random interference such as reflections, occlusions, and other strong interference signals unique to industrial environments. The combined effects of these noise components degrade the quality of the acquired data Y, affecting subsequent surface quality inspection results.

[0059] Therefore, the goal of this embodiment is to use tensor restoration technology to process the noise-affected tensor Y into a clean target tensor X. In other words, clean point cloud information is extracted from the noise, thereby more accurately describing the surface morphology and detailed features of automotive parts and providing high-quality data support for quality inspection.

[0060] The present invention constructs a high-dimensional space representation method through third-order tensor modeling, which effectively improves the expression ability of noise and details in point cloud data, so that the noise components can be removed more accurately in the subsequent restoration processing.

[0061] Step 2: Non-local Deep Tensor Ring Decomposition Model

[0062] To achieve efficient and accurate restoration of point cloud data, this embodiment uses a non-local deep tensor ring decomposition model. This model decomposes the high-dimensional representation of the tensor to extract the core structural features of the data, while effectively suppressing complex noise and preserving high details.

[0063] First, the point cloud data of the automobile parts surface is represented as a high-dimensional tensor X. The structural relationship of the point cloud data is constructed through a series of cyclic multilinear products of third-order tensor ring factors. In this structural relationship, the core features of the point cloud data are efficiently represented through multi-level decomposition. Specifically, through tensor ring decomposition, the tensor X is represented as X = Φ(G), where G = {G (1) ,G (1) ,…,G (N)} is a sequence of third-order core tensors, representing the local features and multi-dimensional geometric structure relationships in the data. Each core tensor Defines the characteristic components of point cloud data in a specific dimension, where r n-1 and rn The rank of the tensor ring is a parameter determined by the specific structural characteristics of the data.

[0064] The core purpose of tensor ring decomposition is to compactly represent the original point cloud data as the product of multiple low-rank tensors. This compact representation can adaptively adjust the weights between different dimensions to a certain extent, allowing the model to adapt to diverse noise environments and data loss situations. Especially in complex industrial environments, this method can significantly improve the robustness to noise. Even if there are large areas of noise or missing areas in the point cloud data, the decomposition model can still maintain an accurate representation of the real signal through the low-rank structure. In addition, through this ring decomposition method, the model can discriminate between different noise types, thereby achieving a good balance between detail preservation and noise suppression.

[0065] Furthermore, the formula X = Φ(G) describes a tensor decomposition method for high-dimensional point cloud data, where the point cloud data X is represented as a structured representation generated by third-order tensor ring decomposition. This tensor ring decomposition method is an efficient and compact data representation technique that can decompose high-dimensional data into a combination of multiple low-rank tensors, facilitating data feature extraction and noise processing. This decomposition structure can adaptively adjust the weights of different data dimensions, effectively suppressing noise interference and providing robustness to missing data.

[0066] In this formula, X represents the restored point cloud data tensor. Ideally, it is a high-dimensional structured representation of noise-free point cloud data that can accurately preserve the geometric shape and texture characteristics of the surface of automotive parts. Φ(G) is an expression generated by tensor ring decomposition, representing the high-dimensional structure of point cloud data, which is composed of multiple third-order tensor factors G (1) ,G (2) ,…,G (1) The core tensor sequence G is the key to describing the local features and multi-dimensional geometric structure of point cloud data. Each core tensor G (n) The dimension is r n-1 ×I n ×r n Here, r n-1 and r n is the rank of the tensor ring, which determines the compactness of the decomposition, and I n It is the feature number of the nth dimension, which is used to reflect the information complexity of this dimension.

[0067] Tensor ring decomposition is an efficient multidimensional data decomposition method, and its specific expression is In this expression, Trace is the matrix trace operation, which is used to reduce the multilinear product results between core tensors to scalars, and G(n) (i n ) is the core tensor G (n) In the i n Through this structure, the overall features of the point cloud data can be compactly represented as a combination of low-rank tensors, making data storage and calculation more efficient while maintaining high fidelity to the geometric structure and texture features of the original data.

[0068] This decomposition structure has multiple functions. First, it can adaptively adjust the weights between different dimensions, making the decomposition result better adapted to the characteristic distribution of point cloud data. Second, the decomposition model can effectively separate and process different types of noise, such as sparse noise and Gaussian noise, to improve the quality of point cloud data. Furthermore, even if part of the point cloud data is missing, this decomposition method can still accurately represent the overall geometric features through its compact structure, thus having strong robustness. Finally, by reducing data redundancy through low-rank decomposition, the efficiency of data processing and the economic efficiency of storage are further improved.

[0069] The specific meanings of the parameters can be further clarified: X is the restored point cloud data tensor, which represents the high-dimensional data result generated during the restoration process; G is the core parameter sequence of tensor decomposition, which is used to compactly represent the features of point cloud data; G (n) It is the basic unit of decomposition, and each core tensor captures the features of a specific dimension.

[0070] By combining these parameters with the tensor ring decomposition formula, efficient restoration of automotive component surface point cloud data can be achieved. This decomposition method demonstrates strong adaptability, both in noise suppression and geometric feature preservation, providing high-quality data support for subsequent quality inspection tasks.

[0071] Step 3: Optimize the model definition

[0072] To achieve efficient restoration of point cloud data, this example constructs an optimization model that integrates multiple fidelity terms and balance parameters to ensure an optimal balance between noise suppression and detail preservation. Specifically, the goal of this optimization model is to minimize the error between the restored data and the acquired data while suppressing noise and preserving the geometric details and texture features in the data. The model is defined as follows:

[0073]

[0074] This optimization model is used to restore point cloud image data through mathematical optimization methods. It addresses noise in the acquired data by comprehensively considering data fidelity, regularization constraints, and noise reduction. The model's goal is to efficiently process the acquired data and generate high-quality restored images, providing reliable support for subsequent surface quality inspections.

[0075] This optimization goal is to find the best restoration result of the point cloud image by comprehensively optimizing the restoration data X, sparse noise S and tensor decomposition core parameters G. is the fidelity term, which measures the error between the restored data X and the collected data Y, while taking into account the impact of sparse noise S on the collected data. This fidelity constraint ensures that the restored point cloud data X is as close as possible to the original collected data Y, thereby preserving the core information of the collected data.

[0076] Item 2 is a regularization term that constrains the difference between the restored data X and the tensor ring decomposition result Φ(G). This regularization constraint ensures that the restored data X conforms to the structural characteristics of the tensor decomposition, thereby improving the restored data's geometric consistency and texture detail preservation. The introduction of regularization effectively improves the applicability of the restored data in complex environments, resulting in more accurate restoration results.

[0077] Item 3 It is a regularization constraint on the sparse noise S, aiming to ensure the sparsity of S and thus suppress the influence of unstructured noise. This constraint prevents the model from overfitting to random interference while preserving the key features of the data.

[0078] The fourth term g(X) is a non-local denoising function, which is used to extract non-local features in the restored data X. Through this function, the model can further optimize the smoothness of the restored data and enhance geometric consistency while suppressing local noise, thereby improving the overall quality of the point cloud image.

[0079] The parameter Y represents the collected point cloud data, which is the original data containing noise and serves as the input of the model. The parameter X is the restored point cloud data. The optimization goal is to generate a noise-free, high-quality point cloud image through model processing. The sparse noise S represents the non-structural noise in the collected data, such as random interference such as occlusion and reflection. The tensor decomposition core parameter G is an important component of the model. The core tensor sequence generated by tensor ring decomposition is used to describe the geometric and texture characteristics of the point cloud data. The tensor decomposition result Φ(G) is a structured expression of the point cloud data, ensuring that the restored data conforms to the essential properties of the data.

[0080] The weight parameters β1, β2, and μ in the model determine the relative importance of each constraint. Parameter β1 controls the degree of match between the restored data X and the tensor decomposition result Φ(G), parameter β2 adjusts the regularization strength of the sparse noise S, and parameter μ influences the balance between noise reduction and detail preservation in the non-local denoising function g(X).

[0081] The square of the Froben i us norm The Euclidean distance between matrices or tensors is a commonly used error metric. By using the norm, the model can intuitively quantify the differences between data and provide a clear target for optimization.

[0082] In summary, this optimization model achieves a comprehensive balance between data fidelity, geometric consistency, and noise reduction through the synergistic effects of a fidelity term, a regularization term, a sparse noise term, and a nonlocal noise reduction term. Under the control of penalty parameters, the model effectively generates high-quality, noise-free point cloud data, providing reliable technical support for subsequent surface quality inspection. This optimization model significantly improves the accuracy of point cloud image restoration and enhances robustness in complex industrial environments.

[0083] In order to further improve the restoration accuracy, this embodiment also adopts a deep convolutional neural network f θ (X0) performs feature extraction and data initialization. Deep convolutional neural networks, leveraging their powerful feature extraction capabilities, can effectively identify and preserve image edges and texture details, helping to construct clearer restored images. During the initialization phase, the model input is a randomly initialized tensor X0. After multiple layers of convolution operations, the tensor's gradient features are extracted. The convolution process enhances image detail by identifying changes in image edges.

[0084] During the feature extraction process, the model uses a convolutional network to generate a gradient image Z and calculates the threshold by averaging the regional grayscale. Where u1 and u2 represent the average grayscale values ​​of adjacent regions, respectively. Through threshold calculation and feature reconstruction, the model accurately captures local features and edge information in the image, ensuring that the restored point cloud data has high resolution and detailed representation, providing a more reliable data foundation for subsequent quality inspections.

[0085] Step 4: ADMM solves the model

[0086] To effectively solve the above optimization model, this embodiment uses the alternating direction method of multipliers (ADMM) to perform step-by-step optimization of each variable in the model. The ADMM method decomposes the complex optimization problem into a series of relatively simple sub-problems, updates each variable in a step-by-step iterative process, and ultimately converges to the global optimal solution. The specific solution process is as follows:

[0087] Solve for G (1) Sub-problem

[0088] To optimize the tensor G (1) To express , we first need to minimize the following objective function:

[0089]

[0090] in, G (1) The second modal expansion of Then it is the tensor G (2) and G (3) The second mode expansion after multiple linear products. This step can optimize G (1) The core tensor expression of , enables it to better capture and adapt to different noise structures in point cloud data, thereby more effectively improving the quality of restored data.

[0091] Solve for G (2) Sub-problem

[0092] In solving G (2) When updating, the calculation formula is:

[0093]

[0094] This step obtains G by expanding and solving the multi-modal tensor. (2) To further reduce the impact of noise on the data. Through this process, G (2) A more optimized expression is obtained, which enables it to more accurately represent the feature information in point cloud data.

[0095] Solve for G (3) Sub-problem

[0096] For the tensor G (3) The update is calculated as follows:

[0097]

[0098] This step is in G (3) The optimization is performed under the multimodal expansion of the point cloud data to further extract and preserve the geometric structure information of the point cloud data. This optimization process helps the model more accurately restore point cloud details in complex data environments.

[0099] Solve the subproblems of S

[0100] The solution goal of sparse noise S is as follows:

[0101]

[0102] After obtaining the optimal solution, the updated value of S is:

[0103] S (t+1) =shr ink(Y-Φ(G) t ,μ)

[0104] This step effectively removes the noise from S through sparse constraints, making it more consistent with the actual data distribution. During the sparse noise optimization process, the soft threshold function shr ink is introduced to suppress S, enabling the model to effectively reduce the impact of noise while retaining more real image information.

[0105] Solve the subproblem of X

[0106] For the solution of tensor X, it is iterated through the following update formula:

[0107]

[0108] In this step, the iterative update of X gradually approaches the exact restored image. g(X) in the formula is an improved nonlocal denoising function that further suppresses noise, ensuring the clarity and detail integrity of X. This iterative update process helps the model extract high-fidelity restored images in complex noisy environments.

[0109] The parameters are described as follows:

[0110] X (t+1) : The restored point cloud data obtained in the current iteration. The updated point cloud data obtained through this iteration gradually approaches the ideal restoration result without noise.

[0111] Y: The original point cloud data collected. The input data containing noise is the basis of the restoration process.

[0112] S (t+1) : The sparse noise of the current iteration. The latest S value obtained by optimizing the sparse noise subproblem is used to remove non-structural noise components in the point cloud data.

[0113] β1Φ(G (t+1) ): Tensor ring decomposition constraint. Φ(G (t+1) ): The latest iteration of tensor ring decomposition results, representing the structured representation of point cloud data.

[0114] β1: Weight parameter of the tensor ring decomposition constraint, used to adjust the impact of ring decomposition on the restoration result.

[0115] The geometric features of the restored data X are kept consistent with the tensor decomposition structure to ensure that the restoration result conforms to the high-dimensional structure of the data.

[0116] g(X): Non-local denoising function. This non-local denoising constraint imposed on X captures the non-local nature of the data, smoothing out noise while preserving important details. This further optimizes the restored image, ensuring clarity and detail integrity.

[0117] 1+β1+μ: normalization factor.

[0118] 1: The default weight of the collected data Y.

[0119] β1: Weight parameter of the tensor ring decomposition constraint.

[0120] μ: Weight of the non-local denoising function.

[0121] Balance the weights of various constraints to ensure the rationality of the restoration results.

[0122] Solve the subproblem of θ

[0123] In order to further improve the restoration accuracy of the model, the solution goal for the deep network weight θ is:

[0124]

[0125] By optimizing the deep network weights θ, this step further enhances the model's ability to identify and suppress noise, making it more robust in complex data environments. Optimizing the deep network enhances the model's ability to capture edge features and detailed information, thereby improving the accuracy and clarity of the restored image.

[0126] The further explanation of the above solution process is:

[0127] To effectively solve the above optimization model, the present invention uses the alternating direction method of multipliers (ADMM) to perform a step-by-step optimization of each variable in the model. The ADMM method is an iterative optimization algorithm that decomposes a complex optimization problem into a series of simpler subproblems, gradually updating each variable to ultimately approach a global optimal solution. This method effectively balances data fidelity, noise suppression, and geometric consistency, ensuring that the restored point cloud data has higher accuracy and robustness.

[0128] First, optimize the tensor G (1) The objective function is optimized by minimizing the error Update tensor G (1) Here, Represents the tensor G (1) The second modal expansion of is the tensor G (2) and G (3) The second mode expansion result of the multilinear product of . This step optimizes G(1) The core expression ability of point cloud data can more accurately capture the geometric features in point cloud data while reducing noise interference.

[0129] Next, optimize the tensor G (2) The update formula is in, is the second modal expansion of the tensor X, and It's G (3) and G (1) The first mode expansion result of the multilinear performance of . This step optimizes G through multi-mode expansion and solution. (2) The expression method makes it more adaptable to the interference of noise in point cloud data and further improves the quality of restored data.

[0130] Then, for the tensor G (3) Optimize. The update formula is In this process, through the (1) and G (2) The multimodal expansion results of Utilization, optimization of G (3) The value of is used to better preserve the geometric features of the point cloud data. This step further enhances the expressiveness and applicability of the model in complex data environments.

[0131] Then, the expression of sparse noise S is optimized. The objective function is By using the sparse constraint, the update formula is S (t+1) =shr ink(Y-Φ(G) t ,μ), where shr ink is a soft threshold function used to control the sparsity of noise. This step effectively suppresses unstructured noise, making the model more consistent with the actual data distribution while avoiding overfitting of noise.

[0132] When optimizing the tensor X, the update formula is used Here, g(X) is a non-local denoising function that further enhances the smoothness and consistency of the restored data. Through this iterative update, the tensor X is gradually optimized, making it closer to the accurately restored high-quality image data.

[0133] Finally, the deep network weight θ is optimized. The objective function is By optimizing and adjusting the weights of the deep network, the model can better identify and suppress noise, while enhancing its ability to capture edge features and detail information in point cloud data, thereby further improving the accuracy and clarity of the restored image.

[0134] Through this step-by-step optimization, the ADMM method decomposes the complex global optimization problem into multiple independent yet closely related subproblems. Each subproblem addresses the optimization requirements of a specific variable, ensuring high levels of restored data fidelity, noise suppression, and geometric consistency. Ultimately, this method significantly improves the accuracy and robustness of point cloud data restoration, demonstrating exceptional adaptability in complex industrial environments and providing strong technical support for surface quality inspection of automotive parts.

[0135] Through this step-by-step solution, the ADMM method effectively decomposes the original complex optimization problem into multiple manageable subproblems, gradually converging to a global optimal solution. Ultimately, this embodiment achieves effective noise suppression and detail preservation while maintaining data accuracy, providing high-quality restored data for automotive component surface quality inspection.

[0136] Through the above steps, this embodiment achieves high-precision restoration of the surface quality of automotive parts, and shows significant advantages in noise suppression and detail preservation in complex industrial environments.

[0137] The above description is only a preferred embodiment of the present invention. Therefore, any equivalent changes or modifications made according to the structure, characteristics and principles described in the scope of the patent application of the present invention are included in the scope of the patent application of the present invention.

Claims

1. A robust image restoration method for surface quality inspection of automotive parts, characterized by: The following steps are involved: Collecting surface point cloud data of automobile parts containing three-dimensional geometric information, and constructing the surface point cloud data of automobile parts into a high-dimensional structured representation; The point cloud data includes target data and different types of noise components, and a data model for noise separation is established based on the point cloud data; The point cloud data is processed using a multidimensional data decomposition method, and the signal and noise are separated through a noise separation data model. Construct an optimization model that includes fidelity terms, regularization terms, and noise reduction terms to optimize the error between the restored data and the collected data; Initialize and enhance the data through the feature extraction module; An iterative optimization method is used to solve the optimization model, and finally noise-free restored point cloud data is obtained; The point cloud image is restored by optimizing the model, and the optimization model is: in, is the fidelity term, g(X) is the non-local denoising function; the parameter Y represents the collected point cloud data, which is the original data containing noise and serves as the input of the model; the parameter X is the restored point cloud data; the sparse noise S represents the unstructured noise in the collected data; the tensor decomposition result Φ(G) is the structured expression of the point cloud data; the parameter β1 controls the degree of matching between the restored data X and the tensor decomposition result Φ(G), the parameter β2 is used to adjust the regularization strength of the sparse noise S, and the parameter μ affects the balance between the non-local denoising function g(X) in denoising and detail preservation; The optimization model is solved step by step using the alternating direction method of multipliers (ADMM), including the following steps: Solve for G (1) Sub-problem: The optimization objective is in and is the modal expansion of the core tensor; Solve for G (2) Sub-problem: Update Solve for G (3) Sub-problem: Update Solve the subproblem of S: the optimization goal is And update S (t+1) =shrink(Y-Φ(G) t ,μ); Solve the subproblem of X: Update Solve the subproblem of θ: Update Among them: G (1) ,G (2) ,G (3) It is a third-order core tensor; it expresses the local features and geometric structure of point cloud data through tensor ring decomposition; Tensor G (1) The second mode expansion and G (2) , G (3) The multimodal expansion results of fold2: modal contraction operation; The expanded form of the tensor X in different modes; representing the matrices after the first mode, second mode, and third mode of X are expanded respectively; t+1 represents the iterative update; G (t+1) : tensor decomposition result updated in the t+1th iteration; S (t+1) : sparse noise after optimization in the t+1th iteration; X (t+1) : The restored tensor calculated in the t+1th iteration; Y: collected point cloud data; Φ(G) t : tensor decomposition result of the current iteration; The square of the Frobenius norm; S: sparse noise; shrink(x,μ): soft threshold function; μ: threshold parameter of the soft threshold function; β1,β2: penalty parameters of the optimization model; g(X): non-local noise reduction function; f θ (X0): Feature tensor extracted by the deep network; θ: weight parameter of deep network; X0: Initial input tensor.

2. The robust image restoration method for surface quality inspection of automotive parts according to claim 1, characterized in that: Using a third-order tensor Model the point cloud data of the surface of automobile parts and define the collected noise pollution point cloud data as tensor Y, satisfying Y=X+N g +S, where N g Represents Gaussian noise, S represents sparse noise, X represents the target tensor of noise-free point cloud data in an ideal situation, M represents the spatial length of the tensor, N represents the width of the tensor, and P represents the number of channels.

3. The robust image restoration method for automobile parts surface quality inspection according to claim 1, characterized in that: The point cloud data is represented as a high-dimensional target tensor X = Φ(G), where G = {G (1) ,G (2) ,…,G (n) } is a third-order core tensor sequence, and each core tensor The third-order tensor ring decomposition structure is used to adaptively adjust the weights of different dimensions to achieve robust processing of different noise types and data missing situations, where r n-1 and r n represents the rank of the tensor ring; I n is the number of features in the nth dimension.

4. The robust image restoration method for automobile parts surface quality inspection according to claim 1, characterized in that: Using deep convolutional neural network f θ (X0) performs feature extraction and data initialization, where the input is a randomly initialized tensor X0, and the image edge and texture detail information is extracted through convolution operation, and the regional threshold is calculated using grayscale average. Complete image reconstruction; where u1 and u2 represent the grayscale averages of adjacent areas respectively.

Images