A quality inspection method for static pressure forming of oilstones based on image processing

Abstract

The present invention relates to the field of image processing, and particularly to a method for detecting the quality of hydrostatic forming of oilstones based on image processing. The method includes: obtaining an infrared image set in the sintering and cooling stage of the oilstone, performing superpixel segmentation to obtain the superpixel regions of each frame of the image, matching the superpixel regions of adjacent frames and establishing a corresponding relationship, and calculating the temperature anomaly characteristic values of each region. Analyze the temperature change trend of each region, calculate the correlation coefficient, introduce the distance and temperature anomaly weight for weighted fusion to obtain a comprehensive correlation index. Combining the temperature anomaly characteristic value and the comprehensive correlation index, evaluate the possibility value of pore anomalies occurring inside each superpixel region, and determine whether there is abnormal pore distribution inside the oilstone to complete the quality detection. The present invention realizes the accurate detection of internal pore anomalies of the oilstone through superpixel segmentation and matching, combined with temperature anomaly characteristics and comprehensive correlation analysis, and improves the reliability of the quality detection of hydrostatic forming.

Description

Technical Field

[0001] The present invention relates to the field of image processing. In particular, it relates to a method for detecting the quality of hydrostatic forming of oilstones based on image processing. Background Art

[0002] An oilstone is a grinding tool made of abrasives (such as alumina, silicon carbide, etc.) and a binder, which is used to finely polish, deburr or trim the surface of a workpiece through manual or mechanical friction. It is applied in fields such as mechanical manufacturing, bearing processing, and tool edge grinding, and is an indispensable tool in precision machining.

[0003] During the production process of oilstones, the forming process is crucial. Hydrostatic forming is an ideal choice for oilstones with high requirements for density uniformity because it can avoid the density gradient problem of traditional die pressing. However, in the hydrostatically formed oilstone, the pores inside may be crushed due to excessive hydrostatic pressure, or the pores may be too large due to insufficient pressure, resulting in uneven pore size or distribution. These problems will affect the heat dissipation, chip removal or lubrication performance of the oilstone, and thus affect its processing effect and service life. Therefore, it is very crucial to detect the quality of hydrostatic forming of oilstones.

[0004] Traditional detection methods have many limitations when detecting abnormal internal pore distribution in hydrostatically formed oilstones. Non-destructive testing methods such as ultrasonic testing and X-ray testing have low resolution, are difficult to accurately identify small pores, and are easily interfered by the internal structure noise of the oilstone, resulting in insufficient detection accuracy. In addition, the detection speed is slow, it is difficult to adapt to large-scale production, and it is also unable to monitor the quality changes in the production process in real time, making it difficult to adjust process parameters in a timely manner. Moreover, for oilstones with irregular shapes, traditional methods are difficult to comprehensively cover the internal area, and rely highly on the experience of the detection personnel, prone to misjudgment or missed judgment. Summary of the Invention

[0005] To solve the problems of traditional detection methods, such as low resolution, insufficient accuracy, slow speed, inability to monitor in real time, poor adaptability to irregular shapes, and being prone to misjudgment and missed judgment when detecting abnormal internal pore distribution in hydrostatically formed oilstones, the present invention provides solutions in the following aspects.

[0006] A quality inspection method for hydrostatic forming of oilstones based on image processing, comprising: obtaining an infrared image set in the sintering and cooling stage of the oilstone and performing preprocessing; performing superpixel segmentation on each frame of oilstone image in the infrared image set to obtain superpixel regions, and sequentially performing superpixel region matching on adjacent two frames of oilstone images in the infrared image set to establish a corresponding relationship, and calculating temperature anomaly characteristic values of each superpixel region in each frame of oilstone image during the cooling process according to the matching result; based on analyzing the temperature change trend of each superpixel region in different frames of oilstone images, calculating the correlation coefficient between each superpixel region and other superpixel regions, and introducing distance weight and temperature anomaly weight for weighted fusion to obtain a comprehensive correlation index, and then combining with the temperature anomaly characteristic value to evaluate the possibility value of whether there is pore anomaly inside each superpixel region; according to the possibility value of pores appearing in each superpixel region, determining whether there is abnormal pore distribution inside the oilstone, and completing the quality inspection objective of the hydrostatic forming of the oilstone.

[0007] The effects are as follows: By performing superpixel segmentation and matching on the infrared image set, combining the temperature anomaly characteristic value and the comprehensive correlation index, it is possible to accurately evaluate the possibility of pore anomaly inside each superpixel region, effectively discover potential defects inside the oilstone, use superpixel segmentation and matching technology to quickly establish the corresponding relationship of superpixel regions in adjacent frame images, improve the detection efficiency, ensure timely discovery of quality problems, comprehensively consider the temperature anomaly characteristic value and the correlation index, comprehensively evaluate the pore distribution inside the oilstone, provide a more accurate basis for quality inspection, through analyzing the temperature change trend and correlation, and performing weighted fusion by combining distance weight and temperature anomaly weight, make the detection result more persuasive and reliable, timely discover abnormal pore distribution inside the oilstone, help to adjust process parameters in time during the production process, improve the performance and service life of the oilstone, and ensure the stable and reliable quality of the product.

[0008] Preferably, the superpixel region includes:

[0009] Obtaining the gray values of all pixel points in each frame of oilstone image, and calculating the gradient amplitude of each pixel point using the sobel operator;

[0010] Based on the gray value, gradient value and position coordinates of pixel points in each frame of oilstone image, using the SLIC superpixel segmentation algorithm to perform superpixel segmentation on the infrared image set to obtain superpixel regions.

[0011] Preferably, the sequentially performing superpixel region matching on adjacent two frames of oilstone images in the infrared image set includes:

[0012] Taking any frame of the oilstone image as the target frame image, obtaining the coordinates of the central pixel points of the superpixel regions in the target frame image, and calculating the Euclidean distance between the central pixel points of the superpixel regions in the target frame image and the subsequent frame of the oilstone image to match the corresponding regions;

[0013] In response to at least two superpixel regions in the subsequent frame of the oilstone image being matched to the target frame image, the relevant superpixel regions in the target frame image are merged; in response to the superpixel regions in the subsequent frame of the oilstone image having no corresponding regions in the target frame image, the superpixel region in the subsequent frame of the oilstone image that is closest to this superpixel region is searched for and merged.

[0014] The effect is that by matching and merging the superpixel regions, the difference in the number of superpixel regions can be effectively reduced, the matching accuracy can be improved, the computational complexity can be reduced, and the detection efficiency and accuracy can be enhanced.

[0015] Preferably, the temperature anomaly eigenvalue includes:

[0016] Taking any frame of the oilstone image as the target frame image and any superpixel region in the target frame image as the significant region, calculating the ratio between the gray mean value of the pixel points in the significant region of the target frame image and the gray mean value of all pixel points in the target frame image, and taking the absolute difference between 1 and the ratio as the deviation degree of the gray mean value; calculating the ratio between the maximum value of the gray gradient of the pixel points in the significant region of the target frame image and the maximum value of the gray gradient of all pixel points in the target frame image to obtain the maximum gradient ratio; taking the ratio of the standard deviation of the gray values of all pixel points in the significant region of the target frame image to the gray mean value as the coefficient of variation; taking the product of multiplying the deviation degree of the gray mean value, the coefficient of variation, and the maximum gradient ratio as the temperature anomaly eigenvalue of the significant region in the target frame image.

[0017] The effect is that by comprehensively considering the deviation degree of the gray mean value, the coefficient of variation, and the maximum gradient ratio of the significant region, the temperature anomaly characteristics of the significant region are accurately quantified. By calculating the deviation degree of the gray mean value, the gray difference between the significant region and the overall image is reflected; the coefficient of variation measures the dispersion degree of the gray values within the region, revealing the temperature distribution uniformity; the maximum gradient ratio reflects the severity of the temperature change. Multiplying these three together to obtain the temperature anomaly eigenvalue can effectively highlight potential temperature anomaly regions and provide a key basis for subsequent analysis.

[0018] Preferably, the correlation coefficient includes:

[0019] According to the correspondence of superpixel regions, determine the correspondence of the same superpixel region in oilstone images of different frames. Taking any frame of oilstone image as the target frame image and any superpixel region in the target frame image as the significant region, obtain the gray mean value of the significant region in the target frame image, and obtain the gray mean value sequence of the significant region during the entire cooling process;

[0020] Traverse all superpixel regions, and use the Pearson correlation coefficient to calculate the correlation coefficient between the gray mean value sequence of the significant region and the gray mean value sequences of other superpixel regions.

[0021] The effect is that: by determining the correspondence of superpixel regions in different frames, the changes of each region during the cooling process can be accurately tracked. Calculating the gray mean value sequence of the significant region during the entire cooling process can reflect the temperature change trend of this region. Using the Pearson correlation coefficient to calculate the correlation coefficient between the gray mean value sequences of the significant region and other regions can quantify the similarity of temperature changes between different regions, and then evaluate the uniformity and potential anomalies of pore distribution. Regions with lower correlation coefficients may indicate abnormal pore distribution because the temperature change trends of these regions are inconsistent with those of other regions, which helps to identify regions that may have internal defects.

[0022] Preferably, the comprehensive correlation index includes:

[0023] Take the absolute value of the correlation coefficient as the temperature change trend during the cooling process. Taking any frame of oilstone image as the target frame image and any superpixel region in the target frame image as the significant region, take the minimum value of the Euclidean distance between the central pixel points of the significant region and other superpixel regions as the distance weight, and take the mean value of the temperature anomaly characteristic values of other superpixel regions in each frame image as the temperature anomaly weight;

[0024] Take the product of the temperature change trend, the negative value of the distance weight, and the temperature anomaly weight respectively as the contribution value of the correlation coefficient of other superpixel regions to the significant region, and sum the contribution values of the correlation coefficients of all superpixel regions as the comprehensive correlation index of the significant region.

[0025] The effect is that: by calculating the correlation coefficient between the significant region and other regions and performing weighting in combination with the distance weight and the temperature anomaly weight, the correlation of the temperature change trends between different regions can be accurately quantified. The product of the temperature change trend, the negative value of the distance weight, and the temperature anomaly weight is used as the contribution value to highlight potential temperature anomaly regions. Summing the contribution values of the correlation coefficients of all superpixel regions to obtain the comprehensive correlation index of the significant region helps to comprehensively evaluate the consistency of the temperature change trend of the significant region with other regions. Through the comprehensive correlation index, regions that may have internal defects can be more accurately identified, improving the accuracy of static pressure forming quality detection.

[0026] Preferably, the possibility value of the occurrence of pore anomalies includes:

[0027] Taking any frame of the oilstone image as the target frame image, taking any superpixel region in the target frame image as the significant region, obtaining the maximum value of the temperature anomaly feature values of all superpixel regions in the target frame image, calculating the ratio between the comprehensive correlation index of the significant region and the maximum value of the comprehensive correlation indexes of all superpixel regions in the target frame image, and obtaining the relative comprehensive correlation coefficient;

[0028] Normalizing the product between the maximum value of the temperature anomaly feature value and the relative comprehensive correlation coefficient to obtain the possibility value of the occurrence of pore anomalies inside the significant region.

[0029] Its effect is that: by combining the maximum value of the temperature anomaly feature value and the relative comprehensive correlation coefficient and using normalization processing, complex indexes can be converted into intuitive possibility values, accurately quantifying the probability of the occurrence of pore anomalies inside the significant region, and normalizing the result to a fixed range, usually , the higher the value, the greater the possibility of pore anomalies. Considering temperature anomalies and correlations comprehensively, highlighting the pore anomaly regions and improving the detection accuracy.

[0030] Preferably, determining whether there are abnormal pore distributions inside the oilstone includes:

[0031] In response to the possibility value of the occurrence of pore anomalies inside the superpixel region being greater than or equal to the preset pore anomaly threshold, there are abnormal pore distributions inside this region, and the control system makes a mark; conversely, if it is less than the preset pore anomaly threshold, the pore distribution inside this region is uniform and meets the standard.

[0032] Preferably, preprocessing the infrared image set includes:

[0033] Using a filtering algorithm to denoise each frame of the oilstone image in the infrared image set, aligning the infrared image set, and normalizing the background of each frame of the oilstone image in the infrared image set to reduce the influence of background differences on defect detection.

[0034] Preferably, the background normalization processing includes:

[0035] Obtaining the original infrared thermal maps of the oilstone surface in multiple frames before heating, and calculating the average value of the original infrared thermal maps of the oilstone surface in multiple frames to obtain a reference background image. After the oilstone undergoes a high-temperature sintering process, using an infrared detector to obtain the infrared image set of the oilstone surface during the cooling stage at a fixed frequency, and subtracting the reference background image from the infrared image sequence after heating to obtain the actual infrared radiation rise value caused by long-pulse excitation on the oilstone surface, thus completing the background normalization processing.

[0036] Its effects are as follows: Background normalization can effectively eliminate the influence caused by the uneven initial thermal radiation distribution of the oilstone, reduce the interference of background differences on subsequent defect detection, and improve the detection accuracy. After background normalization, the actual infrared radiation rise value caused by long-pulse excitation on the oilstone surface can be obtained, which can more accurately reflect the thermal radiation change of the oilstone during the cooling process and is helpful for the subsequent analysis of abnormal pore distribution inside the oilstone.

[0037] The present invention has the following effects:

[0038] 1. Through superpixel segmentation and matching, combining the temperature anomaly eigenvalue and the comprehensive correlation index, the present invention accurately evaluates the possibility of pore anomalies inside each superpixel region, effectively discovers potential defects inside the oilstone, and ensures the quality of static pressure forming.

[0039] 2. Through superpixel segmentation, matching and correlation analysis, the present invention quickly establishes the corresponding relationship of superpixel regions in adjacent frame images, highlights potential temperature anomaly regions, improves the detection efficiency, and comprehensively considers various factors to enhance the detection accuracy.

[0040] 3. By analyzing the temperature change trend and correlation, and performing weighted fusion by combining the distance weight and the temperature anomaly weight, the present invention makes the detection result more persuasive and reliable, providing a stable basis for the quality detection of oilstone static pressure forming. Description of the Drawings

[0041] Figure 1 It is a flowchart of the method of steps S1 - S4 in a method for detecting the quality of oilstone static pressure forming based on image processing according to an embodiment of the present invention. Specific Embodiments

[0042] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention.

[0043] Refer to Figure 1 , a method for detecting the quality of oilstone static pressure forming based on image processing includes steps S1 - S4, specifically as follows:

[0044] S1: Obtain an infrared image set of the oilstone during the sintering and cooling stage, and perform preprocessing.

[0045] Performing preprocessing on the infrared image set includes:

[0046] For the infrared image set, a filtering algorithm is used to denoise each frame of the oilstone image, the infrared image set is aligned, and the background of each frame of the oilstone image in the infrared image set is normalized to reduce the influence of background differences on defect detection.

[0047] To eliminate the influence caused by the movement of the oilstone sample during the sintering and cooling process, it is necessary to align the infrared image set. The feature point matching algorithm (such as the SIFT algorithm) is used. First, the feature points in each frame of the image are detected, and then the image is transformed and aligned according to the matching relationship of the feature points, so as to ensure the consistency of the position and posture of the oilstone in the infrared image set. The feature point matching algorithm is a well-known technology for those skilled in the art and will not be described in detail.

[0048] The background normalization process includes:

[0049] Obtain the original infrared thermal images of the oilstone surface in multiple frames before heating, and calculate the average value of the original infrared thermal images of the oilstone surface in multiple frames to obtain a reference background image. After the oilstone undergoes the high-temperature sintering process, use an infrared detector to obtain the infrared image set of the oilstone surface during the cooling stage at a fixed frequency. Subtract the reference background image from the infrared image sequence after heating to obtain the actual infrared radiation rise value caused by the long-pulse excitation on the oilstone surface, and complete the background normalization process.

[0050] The purpose of obtaining the reference background image is to obtain a stable and representative background thermal radiation distribution.

[0051] S2: Perform superpixel segmentation on each frame of the oilstone image in the infrared image set to obtain superpixel regions. Sequentially perform superpixel region matching on two adjacent frames of the oilstone image in the infrared image set to establish a corresponding relationship. According to the matching result, calculate the temperature anomaly characteristic value of each superpixel region in each frame of the oilstone image during the cooling process.

[0052] The superpixel region includes:

[0053] Obtain the gray values of all pixel points in each frame of the oilstone image, and use the sobel operator to calculate the gradient amplitude of each pixel point;

[0054] Based on the gray values, gradient values, and position coordinates of the pixel points in each frame of the oilstone image, use the SLIC superpixel segmentation algorithm to perform superpixel segmentation on the infrared image set to obtain superpixel regions.

[0055] Exemplarily, the number of superpixel regions is set to 100, which can be appropriately adjusted according to the image size.

[0056] Performing superpixel region matching includes:

[0057] Taking any frame of the oilstone image as the target frame image, obtaining the coordinates of the central pixel points of the superpixel regions in the target frame image, and calculating the Euclidean distance between the central pixel points of the superpixel regions in the target frame image and the subsequent frame of the oilstone image to match the corresponding regions;

[0058] In response to the superpixel regions of the subsequent frame of the oilstone image of the target frame image matching at least two superpixel regions in the target frame image, the relevant superpixel regions in the target frame image are merged; in response to the superpixel regions of the subsequent frame of the oilstone image of the target frame image matching no corresponding superpixel regions in the target frame image, the superpixel region in the subsequent frame of the oilstone image that is closest to this superpixel region is searched for and merged.

[0059] The temperature anomaly eigenvalue includes:

[0060] Taking any frame of the oilstone image as the target frame image and any superpixel region in the target frame image as the significant region, calculating the ratio between the average gray value of the pixel points in the significant region of the target frame image and the average gray value of all pixel points in the target frame image, and taking the absolute difference between 1 and the ratio as the deviation degree of the average gray value; calculating the ratio between the maximum value of the gray gradient of the pixel points in the significant region of the target frame image and the maximum value of the gray gradients of all pixel points in the target frame image to obtain the maximum gradient ratio; taking the ratio of the standard deviation of the gray values of all pixel points in the significant region of the target frame image to the average gray value as the coefficient of variation; taking the product of multiplying the deviation degree of the average gray value, the coefficient of variation, and the maximum gradient ratio as the temperature anomaly eigenvalue of the significant region in the target frame image.

[0061] It should be noted that in the actual physical process, the temperature distribution on the surface of the oilstone is affected by various factors. The temperature anomaly in a certain region is usually not caused by a single factor, but the result of the combined action of multiple factors. Therefore, through the deviation degree of the average gray value, the coefficient of variation, and the maximum gradient ratio, the deviation degree of the average gray value, the uniformity of the temperature distribution within the region, and the severity of the temperature gradient are comprehensively considered, and the possibility of temperature anomaly in this region can be more comprehensively reflected. Only when all three aspects show anomalies is it more likely that there is a real temperature anomaly in this region.

[0062] Specifically, the temperature anomaly eigenvalue satisfies the following relational expression:

[0063] ;

[0064] In the formula, represents the th temperature anomaly eigenvalue of the th superpixel region in the The grayscale mean value of a superpixel region represents the grayscale mean value of all pixel points in the nth frame of the oilstone image represents the coefficient of variation of the grayscale values of all pixel points in the nth frame of the oilstone image in the mth superpixel region the maximum value of the grayscale gradient of pixel points in the mth superpixel region in the nth frame of the oilstone image represents the maximum value of the grayscale gradient of all pixel points in the nth frame of the oilstone image That is to say is the ratio of the grayscale mean value of the mth superpixel region in the nth frame of the oilstone image to the grayscale mean value of the entire oilstone surface region. The closer the ratio is to 1, the closer the temperature of the mth superpixel region is to the average temperature of the entire oilstone surface region, and the lower the possibility of abnormality

[0065] This may be caused by the difference in heat conduction between this region and the whole. For example, the presence of internal pores will change the heat conduction path, resulting in local temperature deviation from the overall average temperature The coefficient of variation reflects the degree of dispersion of the grayscale values of pixel points in this region The larger the value, the more uneven the temperature distribution in this region, and there may be a large temperature gradient, indicating that the heat dissipation in this region is uneven and the possibility of abnormal internal pore distribution is relatively large It is related to the non-uniformity of the microstructure in the region. If the pore distribution in the region is uneven or there are differences in material properties, it will lead to uneven temperature distribution, thus increasing the coefficient of variation is the ratio of the maximum value of the grayscale gradient of pixel points in the mth superpixel region in the nth frame of the oilstone image to the maximum value of the grayscale gradient of pixel points in the entire oilstone surface region. The larger the grayscale gradient of pixel points in this region, the larger the temperature gradient, the worse the heat dissipation effect, and the greater the possibility of abnormal internal pore distribution It is related to the heat flow distribution in the region. A large temperature gradient may mean that the heat flow is blocked or there are abnormal heat sources / sinks, which is related to abnormal pore distribution or other internal defects

[0066]

[0067]

[0068] ​​​​​​​​S3: Based on the analysis of the temperature change trends of each superpixel region in different frames of the oilstone images, calculate the correlation coefficient between each superpixel region and other superpixel regions, and introduce distance weight and temperature anomaly weight for weighted fusion to obtain a comprehensive correlation index. Then, combined with the temperature anomaly eigenvalue, evaluate the possibility value of whether there is pore anomaly inside each superpixel region.

[0069] The correlation coefficient includes:

[0070] According to the corresponding relationship of the superpixel regions, determine the corresponding relationship of the same superpixel region in different frames of the oilstone images. Taking any frame of the oilstone image as the target frame image and any superpixel region in the target frame image as the significant region, obtain the gray mean value of the significant region in the target frame image, and obtain the gray mean value sequence of the significant region during the entire cooling process;

[0071] Traverse all superpixel regions, and use the Pearson correlation coefficient to calculate the correlation coefficient between the gray mean value sequence of the significant region and the gray mean value sequences of other superpixel regions.

[0072] Further analysis shows that since the heat dissipation effect at the edge of the oilstone is better than that in the central region, the correlation of temperature changes between different superpixel regions on the oilstone surface will vary with the distance. If there is an abnormality in the pore distribution inside a certain superpixel region, then the correlation between the temperature change of this superpixel region and other superpixel regions and other superpixel regions will decrease. To avoid the influence of distance, weight the correlation value of the gray change sequence between regions.

[0073] For regions that are closer, the consistency of the temperature change trend is usually higher, and the confidence level of the correlation value is also higher; conversely, for regions that are farther away, the consistency of the temperature change is lower, and the confidence level of the correlation value is lower. Therefore, when calculating the correlation value, a distance weight is introduced, and a lower weight is given to the correlation value of regions that are farther away.

[0074] If there is an abnormality in the pore distribution inside a certain region, its heat dissipation effect will become worse, and the mean value of the temperature anomaly eigenvalue will be larger. In this case, the correlation between the temperature change trend of this region and other regions may decrease. Therefore, when calculating the correlation value, the mean value of the temperature anomaly eigenvalue is introduced as a weight, and a lower weight is given to the correlation value of regions with a larger mean value of the temperature anomaly eigenvalue.

[0075] The specific weighting steps are as follows:

[0076] The comprehensive correlation index includes:

[0077] Take the absolute value of the correlation coefficient as the temperature change trend during the cooling process. Take any frame of the oilstone image as the target frame image, and take any superpixel region in the target frame image as the significant region. Take the minimum value of the Euclidean distance between the center pixel points of the significant region and other superpixel regions as the distance weight, and take the average value of the temperature anomaly eigenvalues of other superpixel regions in each frame image as the temperature anomaly weight;

[0078] Multiply the temperature change trend, the negative of the distance weight, and the temperature anomaly weight respectively, and take the product as the contribution value of other superpixel regions to the correlation coefficient of the significant region. Sum the contribution values of the correlation coefficients of all superpixel regions as the comprehensive correlation index of the significant region.

[0079] By weighting the correlation values of the gray change sequences between regions, the correlation of the temperature change trends between different regions can be measured more accurately, avoiding the influence of distance and pore anomalies on the correlation analysis, so as to more accurately evaluate the possibility of abnormal pore distribution within each superpixel region.

[0080] It should be noted that since the heat dissipation effect at the edge of the oilstone is usually better than that in the central region, the correlation of the temperature change between different regions will change with the distance. The closer the regions are, the higher the consistency of the temperature change trend usually is, and the higher the confidence level of the correlation value is; on the contrary, the farther the regions are, the lower the consistency of the temperature change is, and the lower the confidence level of the correlation value is. Therefore, when calculating the comprehensive correlation index, the distance weight is introduced to correct the confidence level of the correlation value.

[0081] Furthermore, if there are abnormalities in the pore distribution within a certain superpixel region, its heat dissipation effect will deteriorate, and the average value of the temperature anomaly eigenvalues will be relatively large. In this case, the correlation of the temperature change trend between this region and other regions may decrease. Therefore, when calculating the comprehensive correlation index, the average value of the temperature anomaly eigenvalues is introduced as a weight to correct the correlation value. The larger the average value of the temperature anomaly eigenvalues is, the smaller the correlation contribution value is.

[0082] Specifically, the comprehensive correlation index satisfies the following relational expression:

[0083] ;

[0084] In the formula, represents the comprehensive correlation index of the th superpixel region, represents the number of superpixel regions, represents the th superpixel region and the th superpixel region, the Pearson correlation coefficient value between the gray mean value sequences, represents the The Euclidean distance between the center pixel of the th superpixel region and the center pixel of the th superpixel region, represents the mean value of the temperature anomaly eigenvalue of the th superpixel region in all frame images,

[0085] That is to say, the larger the

[0086] value, the greater the correlation between the temperature change trends of these two regions during the cooling process; represents the minimum value of the Euclidean distance between the center pixel of the th superpixel region and the center pixel of the

[0087] th superpixel region in each frame image. The closer the distance between the two regions, the higher the confidence level of the correlation value; conversely, the farther the distance between the two regions, the lower the consistency of their temperature changes and the lower the confidence level of the correlation value. The larger the value, the greater the possibility of abnormal pore distribution inside the

[0088] th superpixel region, the worse the heat dissipation effect, and the greater the possibility of abnormal pore distribution inside it, the lower the confidence level of its correlation value with the

[0089] th region's gray - scale sequence.

[0090] The possibility value of pore abnormality, including:

[0091] Taking any frame of oil - stone image as the target frame image, taking any superpixel region in the target frame image as the significant region, obtaining the maximum value of the temperature anomaly eigenvalues of all superpixel regions in the target frame image, calculating the ratio between the comprehensive correlation index of the significant region and the maximum value of the comprehensive correlation indices of all superpixel regions in the target frame image to obtain the relative comprehensive correlation coefficient;

[0092] ;

[0093] In the formula, represents the possibility value of pore abnormality inside the th superpixel region, represents the normalization function, represents the maximum - value function, represents the The temperature anomaly eigenvalue of the th superpixel region in the frame honing stone image, represents the comprehensive correlation index of the th superpixel region, represents the maximum value of the comprehensive correlation index among the superpixel regions.

[0094] That is to say, obtaining the upper limit value of the temperature anomaly eigenvalues of all superpixel regions in the target frame image aims to: identify the superpixel region with the most abnormal heat dissipation performance in the target frame image, highlighting the extreme situation of uneven potential pore distribution. It represents the highest level of the temperature anomaly eigenvalues of all superpixel regions in the current frame image, is a key indicator for measuring the abnormal heat dissipation situation, and can be used for subsequent analysis of the possibility of abnormal pore distribution inside the honing stone.

[0095] reflects the relative position of the correlation between the temperature change trend of the th superpixel region and other regions in the overall. The smaller the value, the more inconsistent the temperature change trend of this region is with other regions, which may be caused by abnormal pores resulting in abnormal heat dissipation patterns.

[0096] S4: According to the possibility values of pores appearing in each superpixel region, determine whether there is abnormal pore distribution inside the honing stone, and complete the quality inspection objective of the static pressure forming of the honing stone.

[0097] In response to the possibility value of pore anomaly inside the superpixel region being greater than or equal to the preset pore anomaly threshold, there is abnormal pore distribution inside this region, and the control system makes a mark; conversely, if it is less than the preset pore anomaly threshold, the pore distribution inside this region is uniform and meets the standard.

[0098] Exemplarily, the preset pore anomaly threshold is 0.6, and the implementer can adjust it according to the actual situation.

[0099] It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of this invention patent shall be subject to the appended claims.

Claims

1. A quality inspection method for hydrostatic forming of oilstones based on image processing, characterized in that, Including: Obtain the infrared image set in the sintering and cooling stage of the oilstone, and perform preprocessing; Perform superpixel segmentation on each frame of the oilstone image in the infrared image set to obtain superpixel regions. Sequentially perform superpixel region matching on two adjacent frames of the oilstone image in the infrared image set to establish a corresponding relationship. According to the matching result, calculate the temperature anomaly characteristic value of each superpixel region in each frame of the oilstone image during the cooling process; Based on the analysis of the temperature change trend of each superpixel region in different frames of the oilstone image, calculate the correlation coefficient between each superpixel region and other superpixel regions, and introduce distance weight and temperature anomaly weight for weighted fusion to obtain a comprehensive correlation index. Then, combined with the temperature anomaly characteristic value, evaluate the possibility value of whether there is pore anomaly inside each superpixel region, including: Taking any frame of the oilstone image as the target frame image and any superpixel region in the target frame image as the significant region, obtain the maximum value of the temperature anomaly characteristic values of all superpixel regions in the target frame image, and calculate the ratio between the comprehensive correlation index of the significant region and the maximum value of the comprehensive correlation indexes of all superpixel regions in the target frame image to obtain the relative comprehensive correlation coefficient; Normalize the product between the maximum value of the temperature anomaly characteristic value and the relative comprehensive correlation coefficient to obtain the possibility value of pore anomaly inside the significant region; According to the possibility value of pore appearance in each superpixel region, determine whether there is abnormal pore distribution inside the oilstone, and complete the quality inspection objective of the static pressure forming of the oilstone.

2. The method for detecting the hydrostatic forming quality of an oilstone based on image processing according to claim 1, wherein The superpixel region includes: Obtain the gray values of all pixel points in each frame of the oilstone image, and use the sobel operator to calculate the gradient amplitude of each pixel point; Based on the gray value, gradient value, and position coordinates of the pixel points in each frame of the oilstone image, use the SLIC superpixel segmentation algorithm to perform superpixel segmentation on the infrared image set to obtain superpixel regions.

3. A method for detecting the quality of static pressure forming of oilstones based on image processing according to claim 1, characterized in that, The sequential superpixel region matching of two adjacent frames of the oilstone image in the infrared image set includes: Taking any frame of the oilstone image as the target frame image, obtain the coordinates of the central pixel point of the superpixel region in the target frame image, and calculate the Euclidean distance between the central pixel point of the superpixel region in the target frame image and the central pixel point of the superpixel region in the subsequent frame of the oilstone image to match the corresponding region; in response to the superpixel region of the subsequent frame of the oilstone image matching at least two superpixel regions in the target frame image, merge the relevant superpixel regions in the target frame image; in response to the superpixel region of the subsequent frame of the oilstone image not matching any corresponding region in the target frame image, search for the superpixel region closest to this superpixel region in the subsequent frame of the oilstone image for merging.

4. A method for detecting the quality of hydrostatic forming of oil stones based on image processing according to claim 1, characterized in that, The temperature anomaly characteristic value includes: Taking any frame of the oilstone image as the target frame image, and taking any superpixel region in the target frame image as the significant region, calculate the ratio between the gray mean value of the pixel points in the significant region of the target frame image and the gray mean value of all pixel points in the target frame image, and take the absolute difference between 1 and the ratio as the deviation degree of the gray mean value; calculate the ratio between the maximum value of the gray gradient of the pixel points in the significant region of the target frame image and the maximum value of the gray gradient of all pixel points in the target frame image to obtain the maximum gradient ratio; take the ratio of the standard deviation of the gray values of all pixel points in the significant region of the target frame image to the gray mean value as the coefficient of variation; take the product of multiplying the deviation degree of the gray mean value, the coefficient of variation, and the maximum gradient ratio respectively as the temperature anomaly characteristic value of the significant region in the target frame image.

5. The quality inspection method for hydrostatic forming of oilstones based on image processing according to claim 1, characterized in that, The correlation coefficient includes: According to the corresponding relationship of the superpixel regions, determine the corresponding relationship of the same superpixel region in different frames of the oilstone image. Taking any frame of the oilstone image as the target frame image and taking any superpixel region in the target frame image as the significant region, obtain the gray mean value of the significant region in the target frame image, and obtain the gray mean value sequence of the significant region during the entire cooling process. Traverse all superpixel regions and use the Pearson correlation coefficient to calculate the correlation coefficient between the gray mean value sequence of the significant region and the gray mean value sequences of other superpixel regions.

6. The method for detecting the quality of static pressure forming of oil stones based on image processing according to claim 1, characterized in that, The comprehensive correlation index includes: Taking the absolute value of the correlation coefficient as the temperature change trend during the cooling process. Taking any frame of the oilstone image as the target frame image and taking any superpixel region in the target frame image as the significant region, taking the minimum value of the Euclidean distance between the center pixel points of the significant region and other superpixel regions as the distance weight, and taking the mean value of the temperature anomaly characteristic values of other superpixel regions in each frame of the image as the temperature anomaly weight. Taking the product of multiplying the temperature change trend, the negative value of the distance weight, and the temperature anomaly weight respectively as the contribution value of the correlation coefficient of other superpixel regions to the significant region, and summing the contribution values of the correlation coefficients of all superpixel regions as the comprehensive correlation index of the significant region.

7. A method for detecting the quality of hydrostatic forming of oil stones based on image processing according to claim 1, characterized in that, Determining whether there is abnormal pore distribution inside the oilstone includes: In response to the possibility value of pore anomaly appearing inside the superpixel region being greater than or equal to the preset pore anomaly threshold, there is abnormal pore distribution inside this region, and the control system makes a mark; otherwise, if it is less than the preset pore anomaly threshold, the pore distribution inside this region is uniform and meets the standard.

8. A method for detecting the quality of static pressure forming of oilstones based on image processing according to claim 1, characterized in that, Preprocessing the infrared image set includes: Using a filtering algorithm to denoise each frame of the oilstone image in the infrared image set, aligning the infrared image set, and normalizing the background of each frame of the oilstone image in the infrared image set to reduce the influence of background differences on defect detection.

9. A method for detecting the quality of static pressure forming of oil stones based on image processing according to claim 8, characterized in that, The normalization processing of the background includes: Obtain the original infrared thermal images of multiple frames of the oilstone surface before heating, and calculate the average value of the original infrared thermal images of multiple frames of the oilstone surface to obtain a reference background image. After the oilstone undergoes a high-temperature sintering process, use an infrared detector to obtain a set of infrared images of the oilstone surface during the cooling stage at a fixed frequency. Subtract the reference background image from the sequence of infrared images after heating to obtain the actual infrared radiation rise value of the oilstone surface caused by long-pulse excitation, and complete the background normalization process.

Images