A damage identification system for aluminum grinding equipment

Abstract

The invention relates to a damage identification system for aluminum material grinding equipment, belonging to the field of graphic recognition technology. The system comprises the following steps: extracting multiple initial edge lines in the current image after grinding of the aluminum material, and calculating the damage degree of each initial edge line; when the damage degrees of all the initial edge lines do not meet the damage condition, performing differential extraction between the image after grinding again and the current image after grinding; connecting each differential edge line with the initial edge line with the greatest connection necessity to form a new edge line; calculating the difference between the damage degree of each new edge line and the damage degree of the initial edge line connected thereto; calculating the predicted damage step length according to the damage degree difference corresponding to each new edge line; the invention calculates the current damage degree of the aluminum material according to the initial edge lines generated in the current image after grinding under the current grinding times, and predicts and marks the area where damage may be generated by subsequent grinding.

Description

Technical Field

[0001] The invention belongs to the technical field of pattern recognition, and in particular relates to a damage recognition system for aluminum material grinding equipment. Background Art

[0002] In the aluminum surface treatment scenario, some aluminum surfaces need to be brushed, and brushed scratch defects are easily caused during the brushing process. Once brushed scratch defects are caused, the quality of the aluminum is easily reduced. Therefore, after the aluminum is brushed, it is necessary to identify whether there are brushed scratch defects on the aluminum surface. In the prior art, the damage information of aluminum brushed scratches is marked and trained through a neural network and then identified, which requires high computing power for the equipment. At the same time, the poor quality of samples marked subjectively by humans will affect the learning results of the neural network, reducing the accuracy of the judgment of aluminum brushed scratch damage. Summary of the invention

[0003] The invention provides a damage identification system for aluminum material grinding equipment, which is used to solve the problem that the scratch damage defects on the surface of aluminum material cannot be accurately determined in the prior art.

[0004] The damage identification system for aluminum material grinding equipment of the present invention adopts the following technical scheme:

[0005] The system memory, the processor, and the computer program stored in the memory and executable on the processor, wherein the processor implements the following method steps when executing the computer program:

[0006] Capture the current polished image of the aluminum material, and extract multiple initial edge lines in the current polished image;

[0007] Extract the chain code direction vector of each pixel point on each initial edge line, and extract the damage degree of each initial edge line according to the chain code direction vector of each pixel point on each initial edge line and the total number of pixels on each initial edge line;

[0008] If the damage degree of any initial edge line in the current polished image is greater than the damage degree threshold, a scratch damage warning is directly issued;

[0009] If the damage degrees of all the initial edge lines in the current polished image are less than or equal to the damage degree threshold, the aluminum material is polished again and the image after the polishing is captured, and the image after the polishing is differentially extracted from the current polished image;

[0010] Extract multiple differential edge lines from the differential image, and use the damage degree calculation method of each initial edge line to calculate the damage degree of each differential edge line. When the damage degree of any differential edge line is greater than the damage degree threshold, directly issue a scratch damage warning.

[0011] When the damage degree of all differential edge lines is less than or equal to the damage degree threshold, the necessity of connecting each differential edge line with each initial edge line is calculated according to the position of each differential edge line in the image after re-polishing and the position of each initial edge line in the image after re-polishing;

[0012] Connect each differential edge line with the initial edge line with the greatest connection necessity to form a new edge line; calculate the damage degree of each new edge line according to the damage degree calculation method of each initial edge line;

[0013] When the damage degree of any new edge line meets the damage condition, a scratch damage warning is directly issued; when the damage degree of all new edge lines does not meet the damage condition, the difference between the damage degree of each new edge line and the damage degree of the initial edge line connected to it is calculated;

[0014] The predicted damage step length is calculated according to the difference in damage degree corresponding to each new edge line; after polishing again according to the predicted damage step length corresponding to each new edge line, a warning identification mark is made on the image.

[0015] Furthermore, the step of calculating the necessity of connecting each differential edge line with each initial edge line includes:

[0016] According to the distance between the starting point of each differential edge line and the starting point of each initial edge line, and the distance between the end point of each differential edge line and the starting point of each initial edge line, the spatial distance evaluation index between each differential edge line and each initial edge line is calculated respectively;

[0017] According to the direction vectors of the starting point and the end point of each differential edge line and the direction vectors of the starting point and the end point of each initial edge line, the direction consistency evaluation index between each differential edge line and each initial edge line is calculated respectively;

[0018] According to the spatial distance evaluation index and the direction consistency evaluation index between each differential edge line and each initial edge line, the necessity of connecting each differential edge line with each initial edge line is calculated.

[0019] Furthermore, the steps of respectively calculating the spatial distance evaluation index between each differential edge line and each initial edge line include:

[0020] Select any differential edge line as the target differential edge line, and select any initial edge line as the target initial edge line;

[0021] Calculating a first distance between a starting point of a target differential edge line and a starting point of a target initial edge line, and a second distance between an end point of the target differential edge line and a starting point of the target initial edge line;

[0022] calculating a first difference between the first distance and the second distance;

[0023] Using a first ratio of a preset first parameter to the first difference as a spatial distance evaluation index between a target differential edge line and a target initial edge line;

[0024] According to the calculation method of the spatial distance evaluation index between the target differential edge line and the target initial edge line, the spatial distance evaluation index between each differential edge line and each initial edge line is calculated respectively.

[0025] Furthermore, the steps of respectively calculating the direction consistency evaluation index between each differential edge line and each initial edge line include:

[0026] Calculate the first product between the starting and ending direction vectors of the target differential edge line and the starting chain code direction vector of the target initial edge line, and the second product between the absolute value of the starting and ending direction vectors of the target differential edge line and the absolute value of the ending chain code direction vector of the target initial edge line;

[0027] Calculate the third product between the target differential edge line start and end point direction vectors and the target initial edge line end point chain code direction vector, and the fourth product between the absolute value of the target differential edge line start and end point direction vectors and the absolute value of the target initial edge line end point chain code direction vector;

[0028] Calculating a second ratio of the first product to the second product, and simultaneously calculating a third ratio of the third product to the fourth product;

[0029] The second ratio and the third ratio are added and then normalized to extract the direction consistency evaluation index between the target differential edge line and the target initial edge line;

[0030] According to the calculation method of the directional consistency evaluation index between the target differential edge line and the target initial edge line, the directional consistency evaluation index between each differential edge line and each initial edge line is calculated respectively.

[0031] Furthermore, the step of calculating the necessity of connecting each differential edge line with each initial edge line includes:

[0032] The spatial distance evaluation index and the directional consistency evaluation index between each differential edge line and each initial edge line are added and normalized to serve as the connection necessity between each differential edge line and each initial edge line.

[0033] Furthermore, the calculation formula for the damage degree of each initial edge line is:

[0034]

[0035] in, Indicates The degree of damage to the initial edge line; Indicates The first edge line The chain code direction vector of each pixel; Indicates The first edge line The chain code direction vector of each pixel; Indicates The total number of pixels on the initial edge lines; Indicates the preset first parameter; represents the exponential function; represents the normalization function; Indicates the rounding symbol.

[0036] Furthermore, the damage conditions specifically include:

[0037] When the damage degree of any new edge line is greater than or equal to the damage degree threshold, the damage condition is met; when the damage degree of any new edge line is less than the damage degree threshold, the damage condition is not met.

[0038] Furthermore, the step of calculating the predicted damage step length according to the damage degree difference corresponding to each new edge line includes:

[0039] calculating a second difference between the damage level threshold and the damage level of each new edge line;

[0040] A fourth ratio of the second difference value corresponding to each new edge line to the damage degree difference value is calculated, and the fourth ratio is rounded up as the predicted damage step length corresponding to each new edge line.

[0041] Furthermore, the step of extracting a plurality of initial edge lines in the current polished image includes:

[0042] Extract the spectrum of the current polished image;

[0043] The spectrum is subjected to high-pass filtering to extract a high-pass spectrum, and the high-pass spectrum is subjected to inverse Fourier transform to extract the edge image after the initial polishing;

[0044] An edge line in the edge image after the initial grinding is used as an initial edge line.

[0045] Furthermore, the damage level threshold is set to 0.7.

[0046] The beneficial effects of the present invention are:

[0047] The present invention first extracts a plurality of initial edge lines in the current polished image, then calculates the damage degree of each initial edge line, and uses the chain code direction vector of each pixel point on each initial edge line, according to the chain code direction vector of each pixel point on each initial edge line and the total number of pixels on each initial edge line, extracts the damage degree of each initial edge line, and after calculating the damage degree of each initial edge line, directly performs a scratch damage warning when any initial edge line in the aluminum material is severely damaged, thereby preventing the aluminum material with scratch damage from entering the next polishing, and when all the initial edge lines in the aluminum material are not severely damaged, the aluminum material is polished again to obtain the image after the second polishing; due to the polishing The length and direction of the scratches generated in the process change with the number of polishing times in time sequence. Therefore, it is necessary to accumulate the shape features of the scratches generated by the number of polishing times in time sequence to determine the degree of damage caused by the scratches, and predict the possible serious damage area according to the accumulated change of the scratch damage degree. Therefore, the image after the second polishing is differentially extracted from the current image after polishing, and multiple differential edge lines in the differential image are extracted. When the differential edge line is seriously damaged, an alarm is issued, and processing measures are taken in time to avoid expanding the damage effect of the scratch. The present invention directly issues a scratch damage warning when a seriously damaged scratch is generated by calculating the damage degree of the initial edge line and the differential edge line.

[0048] If the differential edge line does not cause serious scratch damage, the damage step length of each scratch is predicted. Therefore, after extracting multiple differential edge lines in the differential image, the present invention connects each differential edge line with the initial edge line with the greatest connection necessity to form a new edge line; according to the damage degree calculation method of each scratch, the damage degree of each new edge line is calculated; when the damage degrees of all new edge lines do not meet the damage conditions, the difference between the damage degree of each new edge line and the damage degree of the initial edge line connected to it is calculated; the predicted damage step length is calculated according to the damage degree difference corresponding to each new edge line; the predicted damage step length corresponding to each new edge line is marked in the image after re-polishing; the present invention also predicts the damage step length of each scratch in the image after re-polishing by predicting the damage step length, which is convenient for identifying scratch damage in the subsequent polishing process to avoid the problem of poor quality of the finished product. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1It is a schematic diagram of the overall steps of an embodiment of a damage identification system for aluminum grinding equipment of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] An embodiment of a damage identification system for aluminum material grinding equipment of the present invention comprises a system memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the following is achieved: Figure 1 Steps shown:

[0053] S1. Capture the current polished image of the aluminum material, and extract multiple initial edge lines in the current polished image.

[0054] The polishing method targeted by the present invention is the most common mechanical polishing method. The dust capture device should be turned on in the polishing production line to prevent the powder generated by polishing from having a significant impact on the imaging effect; the image after each polishing is captured by the capture device installed in the production line.

[0055] The step of extracting multiple initial edge lines in the current polished image includes: extracting the spectrum of the current polished image; extracting the high-pass spectrum after high-pass filtering, and extracting the edge image after initial polishing by inverse Fourier transforming the high-pass spectrum; taking an edge line in the edge image after initial polishing as an initial edge line. The current polished image is grayed to extract the grayscale image, and the two-dimensional discrete Fourier transform is performed to extract the amplitude spectrum, and then the four quadrants are swapped to extract the centered amplitude spectrum, i.e., the spectrum. After the spectrum is high-pass filtered, an inverse Fourier transform is performed, and the high-pass spectrum is transformed to extract the edge image, and the scratch edge image in the current polished image is extracted.

[0056] In the spectrum graph, the high-pass filter blocks low-frequency components and releases high-frequency components. In the polished image, the scratch edges are high-frequency information. The high-frequency edge information is retained through high-pass filtering, and the spectrum image is converted into an edge image through inverse Fourier transform. The high-pass filtering of the spectrum can fully retain the high-frequency edge information in the image, and then a clear edge image is obtained by binarization after inverse Fourier transform. Compared with conventional edge detection such as the Canny operator, the spectrum image retains more edge information and extracts more edges, avoiding excessive loss of edge information.

[0057] S2. Extract the chain code direction vector of each pixel point on each initial edge line, and extract the damage degree of each initial edge line according to the chain code direction vector of each pixel point on each initial edge line and the total number of pixels on each initial edge line.

[0058] The calculation formula for the damage degree of each initial edge line is:

[0059]

[0060] in, Indicates The degree of damage to the initial edge line; Indicates The first edge line The chain code direction vector of each pixel; Indicates The first edge line The chain code direction vector of each pixel; Indicates The total number of pixels on the initial edge lines; Indicates the preset first parameter; represents the exponential function; represents the normalization function; Represents a rounding symbol, which is set to round down in the embodiment of the present invention.

[0061] In the calculation formula of the damage degree of each initial edge line, The initial edge line The cosine values ​​of the angles between the pixel direction vectors are accumulated and then negatively correlated. The larger the value of this formula, the greater the deviation between the pixel directions, and the less likely the edge is to be straight. For the edge The ratio of the number of pixels of the initial edge line to the number of pixels of the initial edge line specified by the production quality can be set according to the production requirements. The larger the scope of possession, the Exceed When The calculation result of is used to evaluate the edge occupancy range; For the existing normalization function, the data falls into interval, which is convenient for screening harmful edges with high damage degree, by calculating the damage degree of each initial edge line , the degree of damage caused to the edges in the image was evaluated.

[0062] Wire drawing is a process of creating a straight edge by friction to achieve certain surface effects. , initial edge line The consistency of the chain code direction means that the edge and the straight line have the same shape. Therefore, the direction vector angle between the edge pixels is used to judge whether the edge is straight. The chain code judges the chain code direction vector by the position relationship of the pixels. The present invention uses the 8-chain code to judge the chain code direction vector of the pixel. The number of edge pixels represents the length of the edge. Under the production quality requirements, if the edge pixels exceed a certain number, it can be judged that the edge does not meet the production requirements. The angle of the edge direction vector is used to judge whether the scratch is straight; the edge length is used to judge whether the scratch is too long.

[0063] S3. If the damage degree of any initial edge line in the current polished image is greater than the damage degree threshold, a scratch damage warning is directly issued.

[0064] Set the damage threshold according to production accuracy requirements , the initial edge line whose damage degree is greater than the damage degree threshold is regarded as the harmful edge. When the scratch damage warning is given directly.

[0065] S4. If the damage degrees of all the initial edge lines in the current polished image are less than or equal to the damage degree threshold, the aluminum material is polished again and the image after the polishing is captured, and the image after the polishing is differentially extracted from the current polished image to obtain a differential image.

[0066] If the damage degree of all initial edge lines in the current polished image is less than or equal to the damage degree threshold When the aluminum material is polished again, the image after the polishing is captured, and the difference image is extracted between the image after the polishing and the current image after the polishing.

[0067] S5. Extract multiple differential edge lines from the differential image, calculate the damage degree of each differential edge line using the damage degree calculation method of each initial edge line, and directly issue a scratch damage warning when the damage degree of any differential edge line is greater than the damage degree threshold.

[0068] If the damage degree of any differential edge line in the differential image is greater than the damage degree threshold When the scratch damage warning is given directly.

[0069] S6. When the damage degree of all differential edge lines is less than or equal to the damage degree threshold, the necessity of connecting each differential edge line with each initial edge line is calculated according to the position of each differential edge line in the image after re-polishing and the position of each initial edge line in the image after re-polishing.

[0070] The step of calculating the necessity of connecting each differential edge line with each initial edge line includes: calculating the spatial distance evaluation index between each differential edge line and each initial edge line according to the distance between the starting point of each differential edge line and the starting point of each initial edge line, and the distance between the end point of each differential edge line and the starting point of each initial edge line; calculating the directional consistency evaluation index between each differential edge line and each initial edge line according to the direction vector of the starting point and end point of each differential edge line and the direction vector of the starting point and end point of each initial edge line; calculating the necessity of connecting each differential edge line with each initial edge line according to the spatial distance evaluation index and the directional consistency evaluation index between each differential edge line and each initial edge line. It should be noted that the method for determining the starting point and end point of the edge is to use the chain code method in the above steps for determination. The specific method is a technical means of numerical value of those skilled in the art, which will not be repeated here.

[0071] After the aluminum material is polished again, the edge of the image after polishing again is superimposed with the edge of the current polishing image or scratches are added. The spectrum images of the nth polishing and the (n+1)th polishing are differentiated by spectrum difference. The changed scratch parts will be shown in the difference image. Each edge line in the difference image is taken as a difference edge line, and the necessity of connecting each difference edge line with each initial edge line is calculated.

[0072] Differential edge line and initial edge line There are many positional relationships between them, but in the final analysis, it is the differential edge line Expanded initial margin The degree of damage , so by differential edge lines With initial edge line The difference edge line is determined by the position relationship of the pixels between Which initial margin line is enlarged? .

[0073] The steps of respectively calculating the spatial distance evaluation index between each differential edge line and each initial edge line include: selecting any differential edge line as a target differential edge line, and selecting any initial edge line as a target initial edge line; calculating a first distance between the starting point of the target differential edge line and the starting point of the target initial edge line, and a second distance between the end point of the target differential edge line and the starting point of the target initial edge line; calculating a first difference between the first distance and the second distance; using a first ratio of a preset first parameter to the first difference as a spatial distance evaluation index between the target differential edge line and the target initial edge line; and respectively calculating the spatial distance evaluation index between each differential edge line and each initial edge line according to a calculation method for the spatial distance evaluation index between the target differential edge line and the target initial edge line.

[0074] Select The differential edge line is taken as the target differential edge line, and the first The initial edge line is used as the target initial edge line, and the The difference edge line and the The spatial distance evaluation index between the initial edge lines The calculation formula is:

[0075]

[0076] in, Indicates The difference edge line and the The spatial distance evaluation index between the initial edge lines; Indicates The starting point of the differential edge line With Initial edge line starting point The first distance between Indicates The total number of pixels contained in the differential edge lines; Indicates End point of differential edge line With Initial edge line starting point The second distance between Indicates the preset first parameter, The initial number of edge pixels specified for production quality ( Can be set according to production requirements).

[0077] In the calculation formula of the spatial distance evaluation index, Indicates The first and last positions of the differential edge line are Initial edge line starting point The reciprocal of the difference in positional Euclidean distances is used to test the positional relationship between the two edges. The first and last positions of the differential edge line are Initial edge line starting point The reciprocal of the difference between the Euclidean distances of the positions and Weighted, indicating the The longer the differential edge line is, the The difference edge line and the The greater the probability of damage after the first initial edge line is connected. The first and last positions of the differential edge line are The spatial distance between the initial edge lines determines the The differential edge line pair The influence range of the initial edge line.

[0078] The steps of respectively calculating the direction consistency evaluation index between each differential edge line and each initial edge line include: calculating the first product between the starting and ending direction vectors of the target differential edge line and the starting chain code direction vector of the target initial edge line, and the second product between the absolute value of the starting and ending direction vectors of the target differential edge line and the absolute value of the ending chain code direction vector of the target initial edge line; calculating the third product between the starting and ending direction vectors of the target differential edge line and the ending chain code direction vector of the target initial edge line, and the fourth product between the absolute value of the starting and ending direction vectors of the target differential edge line and the absolute value of the ending chain code direction vector of the target initial edge line; calculating the second ratio of the first product to the second product, and calculating the third ratio of the third product to the fourth product; adding the second ratio to the third ratio and performing normalization processing to extract the direction consistency evaluation index between the target differential edge line and the target initial edge line; and calculating the direction consistency evaluation index between each differential edge line and each initial edge line according to the calculation method of the direction consistency evaluation index between the target differential edge line and the target initial edge line.

[0079] Select The differential edge line is taken as the target differential edge line, and the first The initial edge line is used as the target initial edge line, and the The difference edge line and the Directional consistency evaluation index between initial edge lines The calculation formula is:

[0080]

[0081] in, Indicates The difference edge line and the Directional consistency evaluation index between initial edge lines; Indicates the The first pixel of the differential edge line is taken as the starting point and the last pixel is taken as the The pixel point is used as the direction vector of the end point; Indicates The first chain code direction vector of the initial edge line; Indicates The total number of pixels on the initial edge lines; Indicates The last chain direction vector of the initial edge line; Represents an existing normalization function.

[0082] In the The difference edge line and the In the calculation formula of the directional consistency evaluation index between the initial edge lines, the first The direction vector of the first and last differential edge lines is The cosine value of the angle between the first chain code direction vector of the initial edge line and the The direction vector of the first and last differential edge lines is The cosine value of the angle between the last chain code direction vector of the initial edge line, the two cosine values ​​of the angle are added and normalized to the interval [0, 1] by sigmoid. The larger the value of this formula, the smaller the value of the first The difference edge line and the The more consistent the directions of the first edge lines are, the The more likely the difference edge line is The purpose of the scratch is to determine whether the directions of the two edges are consistent through the angle between the pixel direction vectors of the two edges.

[0083] The step of calculating the necessity of connection between each differential edge line and each initial edge line includes: adding the spatial distance evaluation index and the directional consistency evaluation index between each differential edge line and each initial edge line and then normalizing them as the necessity of connection between each differential edge line and each initial edge line.

[0084] Necessity of connecting each differential edge line to each initial edge line The calculation formula is:

[0085]

[0086] in, Indicates the necessity of connecting each differential edge line with each initial edge line; Represents the spatial distance evaluation index between each differential edge line and each initial edge line; Represents the directional consistency evaluation index between each differential edge line and each initial edge line; For the existing normalization function (the range is [-1, 1]), The function is normalized to the interval [-1, 1]. Through normalization, the calculation result not only retains the inspection function of the spatial distance position relationship, but also limits the data range to facilitate the screening of the initial edge lines that need to be connected.

[0087] S7. Connect each differential edge line with the initial edge line with the greatest connection necessity to form a new edge line; calculate the damage degree of each new edge line according to the damage degree calculation method of each initial edge line.

[0088] After calculating the connection necessity of each differential edge line with each initial edge line, each differential edge line is connected to the initial edge line with the greatest connection necessity to form a new edge line. Finally, the damage degree of each new edge line is calculated according to the damage degree calculation method of each initial edge line.

[0089] S8. When the damage degree of any new edge line meets the damage condition, a scratch damage warning is directly issued; when the damage degree of all new edge lines does not meet the damage condition, the difference between the damage degree of each new edge line and the damage degree of the initial edge line connected to it is calculated.

[0090] The damage condition is greater than or equal to the damage degree threshold: when the damage degree of any new edge line is greater than or equal to the damage degree threshold, the damage condition is met; when the damage degree of any new edge line is less than the damage degree threshold, the damage condition is not met.

[0091] Connect each differential edge line with the initial edge line with the greatest connection necessity to form a new edge line, and obtain the new edge line The degree of damage , set the damage level threshold ,Will The new edge line is regarded as a harmful edge. When the damage degree of any new edge line meets the damage condition, a scratch damage warning is directly issued. When the damage degree of all new edge lines does not meet the damage condition, the difference between the damage degree of each new edge line and the damage degree of the initial edge line connected to it is calculated.

[0092] The calculation formula for the difference in damage degree is:

[0093]

[0094] in, Indicates the difference in damage degree; Indicates The degree of damage to the new edge line; Indicates The degree of damage to the initial edge line.

[0095] Through the The extent of damage to the new edge line The difference in damage between the adjacent two sides , can calculate the damage step length after how many times of grinding will cause serious damage ,according to Different damage prediction levels of scratches are divided.

[0096] S9. Calculate the predicted damage step length according to the damage degree difference corresponding to each new edge line; and mark the predicted damage step length corresponding to each new edge line on the image after re-polishing.

[0097] The step of calculating the predicted damage step length based on the damage degree difference corresponding to each new edge line includes: calculating the second difference between the damage degree threshold and the damage degree of each new edge line; calculating the fourth ratio of the second difference corresponding to each new edge line to the damage degree difference, and rounding up the fourth ratio as the predicted damage step length corresponding to each new edge line.

[0098] The calculation formula for the predicted damage step length is:

[0099]

[0100] in, represents the predicted damage step length; Indicates the damage level threshold; Indicates The degree of damage to the new edge line; Indicates the difference in damage degree; express .

[0101] The predicted damage step length represents the New edge line passes After the first grinding, serious damage will occur. According to the characteristics of production implementation, Different values ​​of are assigned different tolerances.

[0102] For example: , indicating the A new edge line will be seriously damaged after two grindings, which is highly dangerous; , indicating the A new edge line will be severely damaged after three grindings, which is a medium risk; , indicating the After four grindings, a new edge line will be seriously damaged, which is a slight danger. After extracting the predicted damage step length corresponding to each new edge line, the image after grinding is marked, and each new edge line is given a false color in the image after grinding according to the corresponding danger level, forming a visual display of the judgment information of the damaged area.

[0103] The present invention provides a damage identification system for aluminum grinding equipment, which calculates the current damage degree of the aluminum material according to the initial edge line generated in the current grinding image under the current grinding times, and predicts and marks the area where damage may be generated in subsequent grinding.

[0104] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A damage identification system for aluminum grinding equipment, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the computer program, the following method steps are implemented: Capture the current polished image of the aluminum material, and extract multiple initial edge lines in the current polished image; Extract the chain code direction vector of each pixel point on each initial edge line, and extract the damage degree of each initial edge line according to the chain code direction vector of each pixel point on each initial edge line and the total number of pixels on each initial edge line; If the damage degree of any initial edge line in the current polished image is greater than the damage degree threshold, a scratch damage warning is directly issued; If the damage degrees of all the initial edge lines in the current polished image are less than or equal to the damage degree threshold, the aluminum material is polished again and the image after the polishing is captured, and the image after the polishing is differentially extracted from the current polished image to obtain a differential image; Extract multiple differential edge lines from the differential image, and use the damage degree calculation method of each initial edge line to calculate the damage degree of each differential edge line. When the damage degree of any differential edge line is greater than the damage degree threshold, directly issue a scratch damage warning. When the damage degree of all differential edge lines is less than or equal to the damage degree threshold, the necessity of connecting each differential edge line with each initial edge line is calculated according to the position of each differential edge line in the image after re-polishing and the position of each initial edge line in the image after re-polishing; Connect each differential edge line with the initial edge line with the greatest connection necessity to form a new edge line; calculate the damage degree of each new edge line according to the damage degree calculation method of each initial edge line; When the damage degree of any new edge line meets the damage condition, a scratch damage warning is directly issued; When the damage degree of all new edge lines does not meet the damage condition, the difference between the damage degree of each new edge line and the damage degree of the initial edge line connected to it is calculated; The predicted damage step length is calculated according to the difference in damage degree corresponding to each new edge line; after polishing again according to the predicted damage step length corresponding to each new edge line, a warning identification mark is made on the image; The calculation formula for the damage degree of each initial edge line is: in, Indicates The degree of damage to the initial edge line; Indicates The first edge line The chain code direction vector of each pixel; Indicates The first edge line The chain code direction vector of each pixel; Indicates The total number of pixels on the initial edge lines; Indicates the preset first parameter; represents the exponential function; represents the normalization function; Indicates the rounding symbol.

2. A damage identification system for aluminum material grinding equipment according to claim 1, It is characterized in that The steps of calculating the necessity of connecting each differential edge line with each initial edge line include: According to the distance between the starting point of each differential edge line and the starting point of each initial edge line, and the distance between the end point of each differential edge line and the starting point of each initial edge line, the spatial distance evaluation index between each differential edge line and each initial edge line is calculated respectively; According to the direction vectors of the starting point and the end point of each differential edge line and the direction vectors of the starting point and the end point of each initial edge line, the direction consistency evaluation index between each differential edge line and each initial edge line is calculated respectively; According to the spatial distance evaluation index and the direction consistency evaluation index between each differential edge line and each initial edge line, the necessity of connecting each differential edge line with each initial edge line is calculated.

3. A damage identification system for aluminum material grinding equipment according to claim 2, It is characterized in that The steps of respectively calculating the spatial distance evaluation index between each differential edge line and each initial edge line include: Select any differential edge line as the target differential edge line, and select any initial edge line as the target initial edge line; Calculating a first distance between a starting point of a target differential edge line and a starting point of a target initial edge line, and a second distance between an end point of the target differential edge line and a starting point of the target initial edge line; calculating a first difference between the first distance and the second distance; Using a first ratio of a preset first parameter to the first difference as a spatial distance evaluation index between a target differential edge line and a target initial edge line; According to the calculation method of the spatial distance evaluation index between the target differential edge line and the target initial edge line, the spatial distance evaluation index between each differential edge line and each initial edge line is calculated respectively.

4. A damage identification system for aluminum material grinding equipment according to claim 3, It is characterized in that The steps of respectively calculating the directional consistency evaluation index between each differential edge line and each initial edge line include: Calculate the first product between the starting and ending direction vectors of the target differential edge line and the starting chain code direction vector of the target initial edge line, and the second product between the absolute value of the starting and ending direction vectors of the target differential edge line and the absolute value of the ending chain code direction vector of the target initial edge line; Calculate the third product between the target differential edge line start and end point direction vectors and the target initial edge line end point chain code direction vector, and the fourth product between the absolute value of the target differential edge line start and end point direction vectors and the absolute value of the target initial edge line end point chain code direction vector; Calculating a second ratio of the first product to the second product, and simultaneously calculating a third ratio of the third product to the fourth product; The second ratio and the third ratio are added and then normalized to extract the direction consistency evaluation index between the target differential edge line and the target initial edge line; According to the calculation method of the directional consistency evaluation index between the target differential edge line and the target initial edge line, the directional consistency evaluation index between each differential edge line and each initial edge line is calculated respectively.

5. A damage identification system for aluminum material grinding equipment according to claim 2, It is characterized in that The steps of calculating the necessity of connecting each differential edge line with each initial edge line include: The spatial distance evaluation index and the directional consistency evaluation index between each differential edge line and each initial edge line are added and normalized to serve as the connection necessity between each differential edge line and each initial edge line.

6. A damage identification system for aluminum material grinding equipment according to claim 1, It is characterized in that Damage conditions include: When the damage degree of any new edge line is greater than or equal to the damage degree threshold, the damage condition is met; when the damage degree of any new edge line is less than the damage degree threshold, the damage condition is not met.

7. The damage identification system for aluminum material grinding equipment according to claim 1, It is characterized in that The steps of calculating the predicted damage step length according to the damage degree difference corresponding to each new edge line include: calculating a second difference between the damage level threshold and the damage level of each new edge line; A fourth ratio of the second difference value corresponding to each new edge line to the damage degree difference value is calculated, and the fourth ratio is rounded up as the predicted damage step length corresponding to each new edge line.

8. The damage identification system for aluminum material grinding equipment according to claim 1, It is characterized in that The steps of extracting multiple initial edge lines in the current polished image include: Extract the spectrum of the current polished image; The spectrum is subjected to high-pass filtering to extract a high-pass spectrum, and the high-pass spectrum is subjected to inverse Fourier transform to extract the edge image after the initial polishing; An edge line in the edge image after the initial grinding is used as an initial edge line.

9. The damage identification system for aluminum material grinding equipment according to claim 1, It is characterized in that The damage level threshold is set to 0.7.

Images