Glass wine bottle surface defect detection method and system

Abstract

The invention provides a glass wine bottle surface defect detection system. The glass wine bottle surface defect detection system comprises the following steps: S1, carrying out image acquisition; s2, a step of image processing; s3, a step of defect identification and classification; the step S1 specifically comprises the following steps: S11: selecting and configuring equipment, selecting an industrial-grade high-definition camera which has the characteristics of high resolution, low noise and high dynamic range according to the size, the shape and the surface characteristics of the glass wine bottle so as to ensure that a clear image is captured, and configuring an LED backlight or annular light source system so as to eliminate shadow and light reflection, so that the glass wine bottle is protected; details on the surface of the wine bottle are highlighted, and brightness and color temperature of the light source are adjusted according to scenes; according to the method, the problems of low efficiency and low detection accuracy caused by experience judgment or manual visual observation during existing glass wine bottle detection are well solved, and the intelligent detection method based on visual observation is provided to improve the detection efficiency and accuracy.

Description

Technical Field

[0001] The invention specifically relates to a method and system for detecting surface defects of glass wine bottles. Background Art

[0002] At present, glass wine bottle manufacturers lack standardized, systematic and scientific testing processes and methods when testing wine bottle products. Most of them use empirical judgment or manual intuitive observation. This method is not only inefficient but also has low detection accuracy. Therefore, an intelligent detection method based on visual observation is proposed to solve the above problems. Summary of the invention

[0003] The purpose of the present invention is to provide a glass bottle surface defect detection system in view of the deficiencies in the prior art, and the glass bottle surface defect detection system can well solve the above-mentioned problems.

[0004] In order to achieve the above requirements, the technical solution adopted by the present invention is: to provide a method for detecting surface defects of glass bottles, including the following steps: S1: the step of performing image acquisition; S2: the step of performing image processing; S3: the step of defect identification and classification; the step S1 specifically includes: S11: equipment selection and configuration, according to the size, shape and surface characteristics of the glass bottle, select an industrial-grade high-definition camera, the camera has high resolution, low noise, and high dynamic range characteristics to ensure that a clear image is captured, and an LED backlight or an annular light source system is configured to eliminate shadows and reflections and highlight the details of the bottle surface. The brightness and color temperature of the light source are adjusted according to the scene; S12: installation and debugging, the camera and the light source system are installed on the conveyor belt or rotating table on the production line, and the distance and angle between the camera and the bottle are adjusted to capture Get an image of the entire wine bottle surface, and debug the camera focal length, exposure time, and white balance parameters to obtain the best image quality; S13: Trigger mechanism setting, set the trigger mechanism through a photoelectric sensor, encoder, or image recognition algorithm. When the wine bottle enters the shooting area, the trigger mechanism will send a signal to the camera to start shooting; S14: Real-time preview and adjustment, before formal acquisition, perform a real-time preview to check the image quality, observe through the preview screen whether the details of the wine bottle surface are clear, whether the light source is uniform, and whether there are shadows or reflections, and adjust the parameters of the camera and light source according to the preview results; S15: Image saving and transmission, save the collected images to the local hard disk or transmit them to the server via the network, and record the shooting time and bottle number information of each image when saving the image for subsequent tracing and analysis.

[0005] The advantages of this glass bottle surface defect detection system are as follows:

[0006] This method effectively solves the problem of low efficiency and low detection accuracy caused by the existing glass bottle detection methods using empirical judgment or manual intuitive observation, and proposes an intelligent detection method based on visual observation to improve detection efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The same reference numerals are used in these drawings to represent the same or similar parts. The exemplary embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0008] Figure 1 The figure schematically shows a flow chart of a method for detecting surface defects of a glass bottle according to an embodiment of the present application. DETAILED DESCRIPTION

[0009] In order to make the objectives, technical solutions and advantages of the present application more clear, the present application is further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0010] In the following description, references to "one embodiment", "an embodiment", "an example", "an example", etc. indicate that the embodiment or example described in this way may include specific features, structures, characteristics, properties, elements or limitations, but not every embodiment or example necessarily includes the specific features, structures, characteristics, properties, elements or limitations. In addition, repeated use of the phrase "according to one embodiment of the present application" may refer to the same embodiment, but does not necessarily refer to the same embodiment.

[0011] For the sake of simplicity, certain technical features well known to those skilled in the art are omitted in the following description.

[0012] According to one embodiment of the present application, the detection method used by the glass bottle surface defect detection system includes the following steps:

[0013] 1. Image acquisition

[0014] Sub-step 1: Equipment selection and configuration

[0015] Instructions: First, select a suitable industrial-grade HD camera based on the size, shape and surface characteristics of the glass wine bottle. The camera should have high resolution, low noise, high dynamic range and other characteristics to ensure that clear images are captured. At the same time, configure a suitable light source system, such as LED backlight or ring light source, to eliminate shadows and reflections and highlight the details of the wine bottle surface. The brightness and color temperature of the light source need to be adjusted according to the specific scene to achieve the best shooting effect.

[0016] Sub-step 2: Installation and debugging

[0017] Description: Install the camera and lighting system at a suitable location on the production line, usually on a conveyor belt or a rotating table. Ensure that the camera is at a proper distance and angle from the bottle to capture the image of the entire bottle surface. Debug the camera through the software interface, including adjusting parameters such as focal length, exposure time, and white balance to obtain the best image quality.

[0018] Sub-step 3: Trigger mechanism setting

[0019] Note: To achieve automated acquisition, a trigger mechanism needs to be set up. This can be achieved through photoelectric sensors, encoders, or image recognition algorithms. When the bottle enters the shooting area, the trigger mechanism sends a signal to the camera to start shooting. The accuracy and response speed of the trigger mechanism are crucial to improving acquisition efficiency.

[0020] Sub-step 4: Real-time preview and adjustment

[0021] Note: Before formally capturing, perform a real-time preview to check the image quality. Use the preview screen to observe whether the details on the bottle surface are clear, whether the light source is uniform, and whether there are shadows or reflections. Adjust the parameters of the camera and light source according to the preview results until satisfactory image quality is obtained.

[0022] Sub-step 5: Image saving and transmission

[0023] Note: Save the collected images to the local hard disk or transfer them to the server via the network. When saving images, you need to record the shooting time, bottle number and other information of each image for subsequent tracing and analysis. At the same time, ensure the stability and security of image transmission to avoid data loss or leakage.

[0024] 2. Image Processing

[0025] Sub-step 1: Image preprocessing

[0026] Description: Preprocess the collected original images, including denoising, contrast enhancement, brightness adjustment and other operations. Denoising is to remove random noise points in the image and improve the signal-to-noise ratio of the image; contrast enhancement is to enhance the brightness difference between different areas in the image to make the details clearer; brightness adjustment is to make the overall brightness of the image reach a suitable observation level. These operations help improve the accuracy and efficiency of subsequent processing.

[0027] Sub-step 2: Edge Detection

[0028] Description: Use edge detection algorithm to extract the contour and edge information of the wine bottle surface. The edge is the place where the brightness changes most dramatically in the image, and is usually the potential location of defects. Commonly used edge detection algorithms include Sobel operator, Canny operator, etc. Through edge detection, you can preliminarily screen out areas that may contain defects.

[0029] Sub-step 3: Texture analysis

[0030] Description: Analyze the texture of the bottle surface in detail to identify areas that do not conform to normal texture. Texture analysis can be achieved by calculating the local statistical features of the image through the gray-level co-occurrence matrix, using texture filters or texture recognition algorithms based on deep learning. Texture analysis can further confirm whether the areas screened out in edge detection are indeed defects.

[0031] The specific method of calculating the local statistical features of an image through the gray level co-occurrence matrix is ​​as follows:

[0032] Step 1: Set up the gray-level co-occurrence matrix for multiple directions and distances

[0033] Gray-level co-occurrence matrix value: P d,θ (i, j), represents the co-occurrence frequency of pixel pairs with grayscale values ​​i and j at distance d and direction θ;

[0034] d is the distance between two pixels;

[0035] θ is the angle between two pixels;

[0036] i and j represent the grayscale values ​​of the two pixels respectively, ranging from 0 to L-1, where L is the number of grayscale levels.

[0037] Step 2: Normalize the gray-level co-occurrence matrix and normalize the gray-level co-occurrence matrix for each direction and distance:

[0038]

[0039] Step 3: Extract advanced statistical features and combine the gray-level co-occurrence matrices of multiple directions and distances to extract more complex statistical features.

[0040] energy:

[0041]

[0042] Contrast ratio:

[0043]

[0044] entropy:

[0045]

[0046] Homogeneity:

[0047]

[0048] Relevance:

[0049]

[0050] Inverse variance:

[0051]

[0052] Difference Entropy:

[0053]

[0054] Where p(k)=∑ |i-j|＝k P′ d，θ (i, j) is the probability that the grayscale difference is k.

[0055] Comprehensive feature calculation, weighting all direction and distance features to obtain the final feature:

[0056] Feature=∑ d ∑ θ w d，θ Feature d，θ ;

[0057] w d，θ It is the weight corresponding to direction and distance, which can be adjusted according to application requirements.

[0058] In the above formula:

[0059] P d，θ (i, j): The value of the gray-level co-occurrence matrix, which represents the co-occurrence frequency of pixel pairs with gray-level values ​​i and j at distance d and angle θ.

[0060] P′ d，θ (i, j): Normalized gray-level co-occurrence matrix.

[0061] i, j: Grayscale value of the pixel, ranging from 0 to L-1.

[0062] d: distance between pixels.

[0063] θ: Angle between pixels.

[0064] μ i , μ i : The mean of gray levels i and j.

[0065] σ i , σ j : The standard deviation of gray levels i and j.

[0066] p(k): The probability that the grayscale difference is k.

[0067] w d，θ : The weight under distance d and angle θ.

[0068] Sub-step 4: Color recognition

[0069] Description: For defects with abnormal colors (such as stains), color recognition is required. First, the image is converted from the RGB color space to the HSV color space, which is more suitable for color analysis. Then, the color segmentation algorithm is used to segment the different color areas in the image. Finally, the defects with abnormal colors are identified by comparing the segmented color areas with the preset color templates or samples in the database.

[0070] The method of using the color segmentation algorithm to segment different color areas in an image is as follows:

[0071] (1) Convert the image from RGB color space to HSV color space. The advantage of HSV color space is that it separates color information into three independent components: hue, saturation, and value, making it easier to perform color segmentation. The conversion formula is as follows:

[0072]

[0073] Among them, the calculation of hue θ\thetaθ is as follows:

[0074]

[0075] in:

[0076] R, G, B are the values ​​of the red, green, and blue channels respectively.

[0077] V is the maximum brightness.

[0078] θ is the hue (range 0 to 360 degrees).

[0079] S is the saturation (range 0 to 1)

[0080] (2) Define the color range

[0081] Define the color range to be segmented. In HSV color space, the color range can be specified more intuitively:

[0082] Lower = (Hmin, Smin, Vmin);

[0083] Upper = (Hmax, Smax, Vmax);

[0084] in:

[0085] Hmin and Hmax are the lower and upper limits of the hue, respectively.

[0086] Smin and Smax are the lower and upper limits of saturation respectively.

[0087] Vmin and Vmax are the lower and upper limits of brightness respectively.

[0088] (3) Using the defined color range, perform threshold segmentation on the HSV image to generate a binary mask. In the mask, pixels that match the color range are marked as white (255), and other pixels are marked as black (0);

[0089] (4) To remove noise, morphological operations are performed on the mask, including dilation and erosion. This helps fill small holes and remove isolated noise points. The specific calculation formula is as follows:

[0090] Expansion:

[0091] M dilated (i, j) = max (i′，j′)∈kernel M(i+i′, j+j′);

[0092] corrosion:

[0093] M eroded (i, j) = min (i′，j′)∈kernel M(i+i′, j+j′);

[0094] Where: M dilated (i, j) and M eroded (i, j) are the mask images after dilation and erosion, respectively, and kernel is the structural element of the morphological operation.

[0095] (5) Display the original image and the segmented image to facilitate comparison and verification of the segmentation effect.

[0096] Sub-step 5: Feature extraction and fusion

[0097] Description: Fuse the feature information extracted from edge detection and color recognition. Feature fusion can be performed through weighted fusion to obtain a feature vector containing multiple dimensional information for subsequent defect recognition and classification.

[0098] The specific weighted fusion formula is as follows:

[0099] F=w E ·f(E)+w C g(C)+w EC h(E, C)-λR(E, C);

[0100] in:

[0101] F is the fused feature vector;

[0102] f(E) is a nonlinear transformation of edge features, and f(E) = log(1+E);

[0103] g(C) is a nonlinear transformation of color features, and

[0104] h(E,C) is the interaction function between edge features and color features, and h(E,C) = E·C;

[0105] w E 、w C 、w EC They are the weight coefficients of edge features, color features and feature interactions;

[0106] λ is the weight coefficient of the regularization term;

[0107] R(E,C) is the regularization term, and

[0108] 3. Defect Identification and Classification

[0109] Sub-step 1: Feature matching and comparison

[0110] Description: Compare and match the fused feature vector with the preset defect template or samples in the database. The matching algorithm can be based on statistics. By comparing and matching, determine whether there are defects in the image and the type of defects.

[0111] The specific matching algorithm is as follows:

[0112] D combined (F, T) = ɑ·(1-D P (F, T))+β·D E (F, T)+γ·D M (F, T) + λ·R(F, T);

[0113] in:

[0114] D combined (F, T) is the comprehensive matching measure between the feature vector F and the template vector T.

[0115] D P (F, T) is the Pearson correlation coefficient between the feature vector F and the template vector T.

[0116] D E (F, T) is the Euclidean distance between the feature vector F and the template vector T.

[0117] D M(F, T) is the Manhattan distance between the feature vector F and the template vector T.

[0118] α, β, and γ are weight coefficients of the Pearson correlation coefficient, the Euclidean distance, and the Manhattan distance, respectively, satisfying α+β+γ=1.

[0119] λ is the weight coefficient of the regularization term.

[0120] R(F,T) is the regularization term.

[0121] Sub-step 2: Classifier design and training

[0122] Note: If machine learning is used for defect recognition and classification, a suitable classifier needs to be designed and trained. The design of the classifier includes selecting a suitable algorithm (such as support vector machine, neural network, decision tree, etc.), determining input features, setting algorithm parameters, etc. The training process uses labeled training data (i.e. images with known defect types and locations) to optimize the parameters of the classifier so that it can accurately identify different types of defects.

[0123] Sub-step 3: Classification decision and post-processing

[0124] Description: Make classification decisions based on the output of the classifier. If the classifier outputs multiple possible defect types, you need to follow certain rules.

[0125] The above-mentioned embodiments only represent several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the claims.

Claims

1. A method for detecting surface defects of glass bottles, characterized in that: The steps include: S1: step of image acquisition; S2: image processing step; S3: Defect identification and classification steps; The step S1 specifically includes: S11: Equipment selection and configuration. According to the size, shape and surface characteristics of the glass wine bottle, an industrial-grade high-definition camera is selected. The camera has high resolution, low noise and high dynamic range characteristics to ensure the capture of clear images. At the same time, an LED backlight or ring light system is configured to eliminate shadows and reflections and highlight the details of the wine bottle surface. The brightness and color temperature of the light source are adjusted according to the scene. S12: Installation and debugging: install the camera and light source system on the conveyor belt or rotating table on the production line, adjust the distance and angle between the camera and the wine bottle to capture the image of the entire wine bottle surface, and debug the camera focal length, exposure time, and white balance parameters to obtain the best image quality; S13: Trigger mechanism setting, which is implemented by photoelectric sensors, encoders or image recognition algorithms. When the wine bottle enters the shooting area, the trigger mechanism will send a signal to the camera to start shooting; S14: Real-time preview and adjustment. Before formal acquisition, real-time preview is performed to check the image quality. The preview image is used to observe whether the details on the surface of the wine bottle are clear, whether the light source is uniform, and whether there are shadows or reflections. The parameters of the camera and light source are adjusted according to the preview results. S15: Image saving and transmission: the collected images are saved to the local hard disk or transmitted to the server via the network. When saving the images, the shooting time and wine bottle number information of each image are recorded for subsequent tracing and analysis.

2. The method for detecting surface defects of glass bottles according to claim 1, characterized in that: Step S2 specifically includes: S21: Image preprocessing, preprocessing the collected original image, including denoising, contrast enhancement, and brightness adjustment; S22: Edge detection: Use edge detection algorithm to extract the contour and edge information of the wine bottle surface. The edge is the place where the brightness changes most dramatically in the image, and is usually the potential location of defects. Through edge detection, the area that may contain defects is preliminarily screened out; S23: Texture analysis: Analyze the texture of the wine bottle surface to identify areas that do not conform to normal textures. Texture analysis is achieved by calculating local statistical features of the image through gray-level co-occurrence matrix, using texture filters or texture recognition algorithms based on deep learning. Texture analysis is used to further confirm whether the areas screened out in edge detection are indeed defects. S24: Color recognition: For defects with abnormal colors, color recognition is required. First, the image is converted from the RGB color space to the HSV color space that is more suitable for color analysis. Then, the color segmentation algorithm is used to segment different color areas in the image. Finally, the color abnormality defects are identified by comparing the segmented color areas with the preset color template or samples in the database. S25: Feature extraction and fusion, fusing the feature information extracted from the edge detection and color recognition steps. Feature fusion can be performed through weighted fusion to obtain a feature vector containing multiple dimensional information for subsequent defect recognition and classification.

3. The method for detecting surface defects of glass bottles according to claim 2, characterized in that: The specific method of calculating the local statistical features of the image by using the gray level co-occurrence matrix in step S23 is as follows: S231: Setting a multi-directional and multi-distance gray level co-occurrence matrix; Gray-level co-occurrence matrix value: P d，θ (i, j), represents the co-occurrence frequency of the pixel pair with grayscale values ​​i and j at distance d and direction θ; d is the distance between two pixels; θ is the angle between two pixels; i and j represent the grayscale values ​​of the two pixels, ranging from 0 to L-1, where L is the number of grayscale levels; S242: Normalize the gray-level co-occurrence matrix, and normalize the gray-level co-occurrence matrix of each direction and distance: S233: extracting advanced statistical features, combining gray-level co-occurrence matrices of multiple directions and distances, and extracting more complex statistical features; energy: Contrast ratio: entropy: Homogeneity: Relevance: Inverse variance: Difference Entropy: Where p(k)=∑ |i-j|＝k P′ d，θ (i, j) is the probability that the grayscale difference is k; Comprehensive feature calculation, weighting all direction and distance features to obtain the final feature: Featuer=∑ d ∑ θ w d，θ ·Feature d，θ ; w d，θ It is the weight corresponding to direction and distance, which is adjusted according to application requirements; In the above formula: P d，θ (i, j): the value of the gray-level co-occurrence matrix, which indicates the co-occurrence frequency of pixel pairs with gray-level values ​​i and j at distance d and angle θ; P′ d，θ (i, j): normalized gray-level co-occurrence matrix; i, j: grayscale value of the pixel, ranging from 0 to L-1; d: distance between pixels; θ: angle between pixels; μ i ,μ j : The mean of gray levels i and j; σ i ,σ j : standard deviation of gray levels i and j; p(k): probability that the grayscale difference is k; w d,θ : The weight under distance d and angle θ.

4. The method for detecting surface defects of glass bottles according to claim 2, characterized in that: The method of using the color segmentation algorithm to segment different color areas in the image in step S23 is as follows: S241: The conversion formula for converting an image from RGB color space to HSV color space is as follows: Among them, the calculation of hue θ\thetaθ is as follows: in: R, G, and B are the values ​​of the red, green, and blue channels respectively; V is the maximum brightness; θ is the hue, ranging from 0 to 360 degrees; S is saturation, ranging from 0 to 1; S242: Define color range, define the color range to be segmented, in HSV color space: Lower = (Hmin, Smin, Vmin); Upper = (Hmax, Smax, Vmax); in: Hmin and Hmax are the lower and upper limits of the hue, respectively; Smin and Smax are the lower and upper limits of saturation, respectively; Vmin and Vmax are the lower and upper limits of brightness, respectively; S243: Using the defined color range, threshold segmentation is performed on the HSV image to generate a binary mask, where pixels in the mask that meet the color range are marked as white, and other pixels are marked as black; S244: Perform morphological operations on the mask, including dilation and erosion. The specific calculation formula is as follows: Expansion: M dilated (i,j)=max (i′,j′)∈kernel M(i+i′,j+j′); corrosion: M eroded (i,j)=min (i′,j′)∈kernel M(i+i′,j+j′); Where: M dilated (i,j) and M eroded (i, j) are the mask images after dilation and erosion, respectively, and kernel is the structural element of the morphological operation; S245: Display the original image and the segmented image to facilitate comparison and verification of the segmentation effect.

5. The method for detecting surface defects of glass bottles according to claim 2, characterized in that: The specific weighted fusion formula in step S24 is as follows: F=ω E ·f(E)+ω C ·g(C)+ω EC ·h(E,C)-λR(E,C); in: F is the fused feature vector; f(E) is a nonlinear transformation of edge features, and f(E) = log(1+E); g(C) is a nonlinear transformation of color features, and h(E,C) is the interaction function between edge features and color features, and h(E,C) = E·C; w E 、w C , wE C They are the weight coefficients of edge features, color features and feature interactions; λ is the weight coefficient of the regularization term; R(E,C) is the regularization term, and 6. The method for detecting surface defects of glass bottles according to claim 1, characterized in that: The step S3 specifically includes: S31: Feature matching and comparison: comparing and matching the fused feature vector with the preset defect template or samples in the database. The matching algorithm is based on statistics. By comparing and matching, it is determined whether there are defects in the image and the type of defects. S32: Classifier design and training, using machine learning methods to identify and classify defects, designing and training a suitable classifier. The design of the classifier includes selecting the algorithm, determining the input features, and setting the algorithm parameters. The training process uses labeled training data, that is, images with known defect types and locations, to optimize the classifier parameters so that it can accurately identify different types of defects; S33: Classification decision and post-processing, making classification decisions based on the output results of the classifier.

7. The method for detecting surface defects of glass bottles according to claim 6, characterized in that: The specific matching algorithm in step S31 is as follows: D combined (F,T)=α·(1-D P (F,T))+β·D E (F,T)+γ·D M (F,T)+λ·R(F,T); in: D combined (F,T) is the comprehensive matching measure between the feature vector F and the template vector T; D P (F,T) is the Pearson correlation coefficient between the feature vector F and the template vector T; D E (F,T) is the Euclidean distance between the feature vector F and the template vector T; D M (F,T) is the Manhattan distance between the feature vector F and the template vector T; α, β, and γ are weight coefficients of Pearson correlation coefficient, Euclidean distance, and Manhattan distance, respectively, satisfying α+β+γ=1; λ is the weight coefficient of the regularization term; R(F,T) is the regularization term.

8. A glass bottle surface defect detection system, characterized by: The system uses any one of the glass bottle surface defect detection methods as described in claims 1 to 7 for detection.

Images