A method and device for inspecting the quality of radiator production

Abstract

The present invention relates to the field of image processing technology, and in particular to a method and device for inspecting the quality of heat sink production. The method first uses a camera to obtain a surface image of one side of the fin of the heat sink to be inspected, performs grayscale processing to obtain a heat sink image, and obtains the stamping deformation coefficient, distortion coefficient and distortion condition of each fin edge based on the heat sink image. Then, based on the distortion coefficient and distortion condition of each fin edge, obtain the distortion deformation coefficient of each fin edge. Based on the distortion deformation coefficient and stamping deformation coefficient of the fin edge, obtain the deformation coefficient of each fin edge. When the deformation coefficient of the fin edge is greater than a preset first threshold, the corresponding fin edge is recorded as a deformed fin edge. The present invention can solve the problem that the manual quality inspection of the heat sink is inefficient in the prior art, and the quality inspection of the heat sink using machine vision is difficult, resulting in poor quality inspection results.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method and device for inspecting the production quality of a radiator. Background Art

[0002] The automobile radiator is an indispensable and important component in the cooling system of the automobile water-cooled engine or power battery. The automobile radiator consists of three parts: the water inlet chamber, the water outlet chamber and the radiator core. The coolant flows in the radiator core, and the air passes outside the radiator. The hot coolant cools down due to the heat dissipation to the air, and the cold air heats up due to the absorption of the heat dissipated by the coolant. In the production process of automobile radiators, some common defects may occur, such as improper operation or process control problems in the production process, or the sheet metal is affected by stress during the stamping process, which leads to uneven deformation of the radiator fins and internal fouling. These conditions will greatly affect the heat dissipation efficiency of the radiator; therefore, in order to ensure the heat dissipation efficiency of the radiator, it is necessary to conduct quality inspection on the completed radiators to screen out unqualified radiators.

[0003] In the prior art, human eyes are generally used to observe the side of the radiator fin to determine whether it is deformed. This method is inefficient and the quality inspection effect is poor due to human subjective factors. In addition, using machine vision to analyze radiator fins is a complex and challenging task. There are a large number of fins inside the radiator, and their shapes and arrangements are complex and diverse, making the analysis work extremely complicated, resulting in poor quality inspection of radiator production by machine vision. Summary of the invention

[0004] In order to solve the problem that the manual quality inspection of radiators in the prior art is inefficient and the machine vision quality inspection of radiators is difficult and has poor results, the purpose of the present invention is to provide a method and device for inspecting the production quality of radiators. The technical solutions adopted are as follows:

[0005] In a first aspect, a method for inspecting the quality of heat sink production is provided, the method comprising:

[0006] Step S1: using a camera to obtain a surface image of one side of the fin of the heat sink to be inspected, and performing grayscale processing to obtain a heat sink image;

[0007] Step S2: Obtain the stamping deformation coefficient of each fin edge according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge;

[0008] Step S3: Obtain the distortion coefficient of each fin edge according to the distance between a fin edge and its nearest fin edge in the heat sink image, the average distance between the fin edge and its nearest fin edge, and the number of pixels to the right and left of the fin edge and its fitting line;

[0009] Step S4: Obtain the distortion of each fin edge according to the absolute value of the difference in fitting loss values ​​between a fin edge and the next fin edge in the distortion direction of the fin edge in the heat sink image, and the absolute value of the difference in distance between a fin edge and the next fin edge in the distortion direction and the adjacent fin edge in the distortion direction;

[0010] Step S5: obtaining the distortion coefficient of each fin edge according to the distortion coefficient and distortion condition of each fin edge;

[0011] Step S6: obtaining the deformation coefficient of each fin edge according to the distortion deformation coefficient and the stamping deformation coefficient of the fin edge;

[0012] Step S7: when the deformation coefficient of the fin edge is greater than a preset first threshold, the corresponding fin edge is recorded as a deformed fin edge.

[0013] Furthermore, after step S1 and before step S2, the method further includes: filtering out the fin edges in the radiator image from all the edges in the radiator image according to the similarity between the fin edges in the radiator image.

[0014] Furthermore, the method of filtering out the fin edges in the radiator image from all the edges in the radiator image based on the similarity between the fin edges in the radiator image is specifically as follows: based on any edge in the radiator image and its similar edges, and the distance between any edge in the radiator image and its similar edges, the fin coefficient of each edge in the radiator image is calculated; when the fin coefficient of an edge in the radiator image is greater than a preset second threshold, the edge is recorded as a fin edge.

[0015] Further, in step S2, the stamping deformation coefficient of each fin edge is obtained based on the mean difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the mean grayscale value on both sides of the turning point of the fin edge;

[0016] The angle between any pixel point on the edge of the fin and the line segment formed by its two adjacent edge pixels on the left and right is recorded as the turning degree of the pixel point on the edge of the fin;

[0017] Compare the turning degree of the pixel point on the edge of the fin with its adjacent pixel points. If the difference between the turning degrees of the pixel point and the pixel points on both sides of it is greater than a third threshold, record it as a turning point.

[0018] The stamping deformation coefficient of each fin edge is obtained according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge.

[0019] Furthermore, when the average difference of the turning degree between the fin edge pixel point and its next pixel point in the heat sink image is larger and the difference of the gray value averages on both sides of the turning point of the fin edge is larger, the stamping deformation coefficient of the corresponding fin edge is larger.

[0020] Furthermore, in step S3, the greater the absolute value of the difference between the distance between a fin edge and its nearest fin edge and the average value of the distance between the fin edge and its nearest fin edge, the greater the distortion coefficient of the fin edge;

[0021] When the ratio between the number of pixels on the right and left sides of the fin edge and its fitting straight line approaches a constant of 1, the distortion coefficient of the fin edge is smaller.

[0022] Furthermore, in step S4, the smaller the absolute value of the difference in fitting loss values ​​between a fin edge and the next fin edge in the twisting direction of the fin edge in the radiator image, and the smaller the absolute value of the difference in distance between a fin edge and the next fin edge in the twisting direction and the adjacent fin edges in the twisting direction, the more serious the distortion of the fin edge.

[0023] Furthermore, in step S5, when the distortion coefficient of the fin edge is larger and the distortion condition is more serious, the distortion deformation coefficient of the fin edge is larger.

[0024] Furthermore, in step S6, when the twisting deformation coefficient of the fin edge is larger and the stamping deformation coefficient is larger, the deformation coefficient of the fin edge is larger.

[0025] In a second aspect, a device for inspecting the quality of heat sinks is provided, the device comprising:

[0026] An image acquisition module is used to acquire a surface image of one side of the fin of the radiator to be inspected through a camera, and obtain a radiator image after grayscale processing;

[0027] A stamping deformation coefficient acquisition module is used to obtain the stamping deformation coefficient of each fin edge according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge;

[0028] A distortion coefficient acquisition module is used to obtain the distortion coefficient of each fin edge according to the distance between a fin edge and its nearest fin edge in the heat sink image, the average distance between the fin edge and its nearest fin edge, and the number of pixels on the right and left sides of the fin edge and its fitting straight line;

[0029] A distortion condition acquisition module is used to acquire the distortion condition of each fin edge according to the absolute value of the difference in fitting loss values ​​between a fin edge and the next fin edge in the distortion direction of the fin edge in the heat sink image, and the absolute value of the difference in distance between a fin edge and the next fin edge in the distortion direction and the adjacent fin edge in the distortion direction;

[0030] A distortion deformation coefficient acquisition module is used to acquire the distortion deformation coefficient of each fin edge according to the distortion coefficient and distortion condition of each fin edge;

[0031] A deformation coefficient acquisition module is used to acquire the deformation coefficient of each fin edge according to the twisting deformation coefficient and the stamping deformation coefficient of the fin edge;

[0032] The alarm module records the corresponding fin edge as a deformed fin edge and issues an unqualified alarm when the deformation coefficient of the fin edge is greater than a preset first threshold.

[0033] The present invention has the following beneficial effects: it can solve the problem that manual quality inspection of radiators in the prior art is inefficient, and that quality inspection of radiators using machine vision is difficult, resulting in poor quality inspection results. First, a camera is used to obtain a surface image of one side of the fin of the radiator to be inspected. Since there is no need to analyze the color space for analysis, the radiator image is obtained after grayscale processing; since the main source of deformation of the radiator during the radiator quality inspection process is caused by the stamping deformation of each fin and the distortion deformation of each fin during assembly and welding, the present invention obtains the deformation coefficient of each fin edge by analyzing the stamping deformation coefficient and distortion deformation coefficient of each fin edge of the radiator image; finally, when the deformation coefficient of the fin edge is greater than the preset first threshold, the corresponding fin edge is recorded as a deformed fin edge. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are 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.

[0035] Figure 1 A flow chart of a method for inspecting good quality of heat sink production provided by one embodiment of the present invention.

[0036] Figure 2 A block diagram of a device for inspecting the quality of heat sinks provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the method and device for inspecting the quality of the production of a radiator proposed by the present invention, its specific implementation method, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. Unless otherwise defined, all technical and scientific terms used in this article have the same meaning as those generally understood by technicians in the technical field of the present invention.

[0038] The scenario targeted by the present invention is: for the radiator that has been produced, since the metal sheet needs to be stamped during the production process of the radiator, the sheet may be deformed due to stress during the stamping process. After the subsequent welding is completed, it enters the quality inspection stage and uses machine vision to check the deformation of the fins of the radiator.

[0039] The specific scheme of the method and device for inspecting the quality of heat sinks provided by the present invention will be described in detail below in conjunction with the accompanying drawings.

[0040] First, see Figure 1 , which shows a flow chart of a method for inspecting the quality of a heat sink according to an embodiment of the present invention, the method comprising the following steps:

[0041] Step S1: Use a camera to obtain a surface image of one side of the fin of the heat sink to be inspected, and obtain a heat sink image after grayscale processing.

[0042] For the finished radiator, after entering the quality inspection stage, the processed radiator fins are placed with one side facing up, and an industrial high-resolution camera is installed above the radiator to obtain the surface image of the radiator to be inspected, and then further analyze whether the radiator fins are deformed. Since the bending and deformation of the radiator fins only need to be analyzed from the edge, and the color space does not need to be analyzed, the radiator surface image is grayed to obtain the radiator surface grayscale image, which is used as the radiator image for subsequent analysis of whether the radiator fins are deformed.

[0043] Since the radiator is welded from several stamped plates, the plates will deform during stamping, which will affect the heat dissipation performance of the radiator. Therefore, the fin edges of the radiator image must be detected to determine whether the fin edges are deformed. There are many edges in the radiator image obtained by the above operation. This is because the radiator not only has cooling fins but also structures such as heat pipes. Edge detection will detect non-fin edges such as heat pipes, which will interfere with fin edge detection. Therefore, it is necessary to filter out the fin edges in the radiator based on the edges in the image, and then perform fin deformation detection based on the obtained fin edges.

[0044] Therefore, after step S1 and before step S2, the method further includes: filtering out the fin edges in the radiator image from all the edges in the radiator image according to the similarity between the fin edges in the radiator image.

[0045] In this embodiment, the fin edges in the radiator image are screened out from all the edges in the radiator image based on the similarity between the fin edges in the radiator image. Specifically, the fin coefficient of each edge in the radiator image is calculated based on any edge in the radiator image and its similar edges, and the distance between any edge in the radiator image and its similar edges. When the fin coefficient of an edge in the radiator image is greater than a preset second threshold, the edge is recorded as a fin edge.

[0046] Specifically, since the fins in the radiator are stamped metal sheets of the same shape and are regularly arranged and then welded, the fins are highly similar and regularly arranged. Therefore, for each radiator edge obtained in the radiator image, the DTW value (Dynamic Time Warping, dynamic time warping algorithm, that is, the optimal path value between the internal edges of the radiator is obtained according to the principle of shortest distance) between the internal edges of each radiator is obtained, and the edge with the smallest DTW value among the edges of each radiator is determined and recorded as a similar edge of the edge; thereby, the fin coefficient of each radiator edge can be obtained.

[0047] In this embodiment, the mathematical formula of the fin coefficient of the ith edge in the constructed heat sink image is as follows:

[0048]

[0049] in, is the index of any edge in the heat sink image, is the number of similar edges to any edge in the radiator image, The radiator image is Strip edge, The radiator image is Similar edges of strip edges, The radiator image is The distance between the edge of a line and its similar edge (i.e., the average distance between all points between two curves in image processing, a well-known technique, which will not be discussed in detail in this embodiment), is the variance of the distances between all edges in the image and their similar edges, is an exponential function with the natural constant e as base. Refers to the DTW value of similar structures or similar paths, that is, the best path value or the minimum path value. Refers to the first The edge and other The minimum DTW value between similar edges of the edge (i.e., the The edge and other The best path value between similar edges of an edge), The radiator image is Fin coefficient of the strip edge.

[0050] In the mathematical formula of the fin coefficient of the ith edge in the above heat sink image, the natural constant e is set to avoid the result being 0. The edge and other Minimum DTW value between similar edges of a strip edge The smaller the value, the higher the similarity between the edge and its similar edges. Since the radiator is composed of the same fins, the edge belongs to the fin edge. Fin coefficient of strip edge is negatively correlated; and for the variance between the distances between an edge in the image and its similar edges The smaller the value, the smaller the distance difference between the edge and other similar edges is. Fin coefficient of strip edge There is a negative correlation, that is, the smaller its value is, the higher the fin coefficient is.

[0051] Furthermore, when the fin coefficient of an edge in the heat sink image is greater than a preset second threshold, the edge is recorded as a fin edge. In this embodiment, the second threshold can be set to 0.8, that is, when the fin coefficient of an edge in the heat sink image is greater than a preset second threshold, the edge is recorded as a fin edge. When 0.8, the edge is recorded as the fin edge.

[0052] It should be further explained that, for the fin edges in the radiator image obtained by the above operation, due to problems with the stamping machine when the fins are crimped, the fin edges are deformed during stamping, resulting in wrinkles and eversion on the fin edges. Therefore, it is necessary to analyze whether the fin edges are smooth, and if the fin edges are not smooth, whether the uneven areas are everted, thereby obtaining the stamping deformation coefficient of each fin edge.

[0053] Step S2: Obtain the stamping deformation coefficient of each fin edge according to the average difference of the turning degree between the fin edge pixel point and its next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge.

[0054] Among them, in step S2, the stamping deformation coefficient of each fin edge is obtained according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the radiator image and the difference of the mean grayscale values ​​on both sides of the turning point of the fin edge. Specifically, the angle between the line segment connecting any pixel point on the fin edge and its two adjacent edge pixels on the left and right is recorded as the turning degree of the fin edge pixel point; the turning degree of the pixel point on the fin edge is compared with its adjacent pixels, and if the difference in turning degree between the pixel point and the pixels on both sides of it is greater than the third threshold, it is recorded as a turning point; the stamping deformation coefficient of each fin edge is obtained according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the radiator image and the difference of the mean grayscale values ​​on both sides of the turning point of the fin edge.

[0055] Specifically, the deformation caused by the stamping of the fin will cause the edge of the fin to deform and wrinkle, so there will be an uneven part on the edge. In order to obtain its unevenness, for any pixel point on the edge of the fin, the angle between the line segment connecting the two adjacent edge pixel points on its left and right is recorded as the turning degree of the edge pixel point, and the unevenness of the edge is obtained according to the difference between the turning degrees of each pixel point; and the turning degree of the pixel point on the edge of the fin is compared with its adjacent pixel points. If the turning degree difference between the pixel point and the pixel points on both sides of it is greater than a third threshold, such as the third threshold may be 0.7, that is, the turning degree difference between the pixel point and the pixel points on both sides of it is greater than 0.7, it is recorded as a turning point; since the deformed turning part may have the edge of the fin flipped out, this part has a large difference in grayscale values ​​on both sides due to the metal reflection of the fin. Therefore, for the turning point, the average grayscale value on both sides of the edge is obtained for comparison, thereby obtaining the stamping deformation coefficient of the fin edge.

[0056] In this embodiment, the constructed radiator image The mathematical formula for the stamping deformation coefficient of the edge of a fin is:

[0057]

[0058] in, is the serial number of any fin edge in the heat sink image, is the serial number of the pixel on the edge of any fin, It is The number of pixels on the edge of a fin, excluding the first and last pixels. The radiator image is The stamping deformation coefficient of the fin edge is It is The first The turning point of each pixel, It is The first The turning point of each pixel, It is The number of turning point differences between adjacent pixels on the edge of a fin, They are the mean gray values ​​on the right and left sides of the edge of the turning point, is a hyperbolic function.

[0059] In the above heat sink image, In the mathematical formula of the stamping deformation coefficient of the fin edge, the average difference between the turning degree of the fin edge pixel point and the next pixel point in the image is The larger the value, the greater the difference in turning degree between the pixels on the edge of the fin, and thus the less smooth the edge is. Therefore, it is positively correlated with the stamping deformation coefficient. And the difference in the mean grayscale value on both sides of the turning point of the fin edge is The larger its value is, the greater the difference in grayscale values ​​on both sides of the turning point of the fin edge is, which indicates that it is more likely that the fin edge is turned out due to stamping deformation, so it is positively correlated with the stamping deformation coefficient.

[0060] Therefore, when the average difference of the turning degree between the fin edge pixel point and its next pixel point in the heat sink image is larger, and the average difference of the grayscale value on both sides of the turning point of the fin edge is larger, the stamping deformation coefficient of the corresponding fin edge is larger.

[0061] After stamping, multiple fins need to be welded to complete a heat sink. Therefore, after analyzing the stamping deformation of the fin edge, the welding process between the fins needs to be considered. During the fin welding process, the fins are squeezed against each other, resulting in inconsistent sparseness of the fins after welding. Some fins have different distances between fins due to squeezing. The fins are twisted and deformed due to the squeezing stress between the fins, and the deformed part of the fins points from the densely squeezed part to the sparse part due to the stress. Therefore, it is necessary to obtain the distortion coefficient of the fins in the image.

[0062] Step S3: Obtain the distortion coefficient of each fin edge according to the distance between a fin edge and its nearest fin edge in the heat sink image, the average distance between the fin edge and its nearest fin edge, and the number of pixels to the right and left of the fin edge and its fitting line.

[0063] Specifically, for each fin edge in the radiator image, the distance between the fin edge closest to it is obtained. Since the fins are arranged regularly under normal circumstances, the minimum distance between the fin edges is similar under normal circumstances. On the other hand, the obtained fin edges are fitted with a straight line using the least squares method. According to the number of pixels on both sides of the fitting line, the side with more pixels is recorded as the distorted side. In this way, the distortion coefficient of the fin edge can be obtained.

[0064] In this embodiment, the least square method is used to fit the obtained fin edge to a straight line. The number of pixels on both sides of the fitting straight line is taken as an example. The number of pixels on the left side is greater than the number of pixels on the right side. The mathematical formula for the distortion coefficient of the edge of a fin is:

[0065]

[0066] in, is the serial number of any fin edge in the heat sink image, is the sequence number of any fin edge and its nearest fin edge, It is The distance between the edge of a fin and the edge of the nearest fin, is the mean distance between the fin edge and its nearest fin edge in the heat sink image, , is the number of pixels on the right and left sides of the fin edge and its fitting line in the image, is a hyperbolic function, Indicates The distortion factor of the fin edge.

[0067] In the heat sink image constructed above In the mathematical formula for the distortion coefficient of the edge of a fin, Indicates The difference between the distance between the edge of the first fin and its nearest fin edge and the average distance between the edge of the fin and its nearest fin edge in the heat sink image. The larger the difference, the greater the distortion of the fin edge. The distortion factor of the fin edge The bigger; It indicates the difference in the number of pixels on both sides of the fitting line. The greater the difference, the greater the distortion of the fin edge. The distortion factor of the fin edge That is, in step S3, when the absolute value of the difference between the distance between a fin edge and its nearest fin edge and the average of the distances between the fin edge and its nearest fin edge is larger, the distortion coefficient of the fin edge is larger; when the ratio between the number of pixels on the right and left sides of the fin edge and its fitting straight line is closer to a constant 1, the distortion coefficient of the fin edge is smaller.

[0068] In addition to considering the distortion coefficient of the fin edge mentioned above, due to the influence of stress, the deformed part of the fin will point from the densely squeezed part to the sparse part, that is, the angle formed by the fin edge points to the sparse fin. The acquired fin edge is fitted with a linear method using the least squares method to obtain its fitting loss value. The difference change of its loss value in the distortion direction of the fin edge and the difference change of the distance between a fin edge and the next fin edge in the distortion direction of the fin edge are combined to evaluate the change of the distortion deformation of the fin edge, and further evaluate the propagation length of the distortion deformation of the fin edge.

[0069] Step S4: Obtain the distortion of each fin edge according to the absolute value of the difference in fitting loss values ​​between a fin edge and its next fin edge in the distortion direction of the fin edge in the radiator image, and the absolute value of the difference in distance between a fin edge and its next fin edge in the distortion direction and the adjacent fin edge in the distortion direction.

[0070] Specifically, in step S4, the smaller the absolute value of the difference in fitting loss values ​​between a fin edge and the next fin edge in the twisting direction of the fin edge in the radiator image, and the smaller the absolute value of the difference in distance between a fin edge and the next fin edge in the twisting direction and the adjacent fin edges in the twisting direction, the more serious the distortion of the fin edge.

[0071] In this embodiment, the constructed radiator image The mathematical formula for the distortion of the edge of a fin is as follows:

[0072]

[0073] in, is the serial number of any fin edge in the direction of its twisting deformation, is the number of fin edges in the direction of the distortion of any fin edge, It is The first fin edge twist direction The absolute value of the difference in fitting loss between the edge of a fin and the edge of the next fin, It is The first fin edge twist direction The absolute value of the difference between the edge of the first fin and the edge of the next fin and the distance between the edge of the next fin and the edge of the next fin, that is, The first fin edge twist direction The absolute value of the difference between the edge of a fin and the edge of the next fin in the twisting direction and the edge of the adjacent fin, It is an exponential function with the natural constant e as the base. The natural constant e is set to avoid the result being 0. Indicates The distortion of the fin edges.

[0074] In the above heat sink image, In the mathematical formula for the distortion of the edge of a fin, The first fin edge twist direction The absolute value of the difference between the edge of a fin and the edge of the next fin in the twist direction and the edge of the adjacent fin With The first fin edge twist direction The absolute value of the difference in fitting loss between the edge of a fin and the edge of the next fin The smaller the two values ​​are, the slower the change of the fin's extrusion and twisting deformation is, and the longer the propagation length of the twisting deformation is, so the greater the twisting condition of the part where it is located.

[0075] Step S5: Obtain the distortion coefficient of each fin edge according to the distortion coefficient and distortion condition of each fin edge.

[0076] Specifically, in step S5, when the distortion coefficient of the fin edge is larger and the distortion condition is more serious, the distortion deformation coefficient of the fin edge is larger.

[0077] In this embodiment, the constructed radiator image The mathematical formula for the distortion coefficient of the edge of a fin is as follows:

[0078]

[0079] in, Indicates The distortion coefficient of the fin edge, Indicates The distortion of the fin edge, Indicates The distortion coefficient of the fin edge.

[0080] In the above heat sink image, In the mathematical formula of the distortion coefficient of the fin edge, the distortion coefficient of the fin edge and the distortion of the fin edge are positively correlated with the distortion coefficient of the fin edge. The larger the distortion coefficient of the fin edge and the more severe the distortion, the larger the distortion coefficient of the fin edge.

[0081] Step S6: Obtain the deformation coefficient of each fin edge according to the twisting deformation coefficient and the stamping deformation coefficient of the fin edge.

[0082] Specifically, in step S6, when the twisting deformation coefficient of the fin edge is larger and the stamping deformation coefficient is larger, the deformation coefficient of the fin edge is larger.

[0083] In this embodiment, the constructed radiator image The mathematical formula for the deformation coefficient of the edge of a fin is as follows:

[0084]

[0085] in, Indicates the first The distortion coefficient of the fin edge is The radiator image is The stamping deformation coefficient of the fin edge is Indicates the first The deformation coefficient of the fin edge.

[0086] In the above heat sink image, In the mathematical formula for the deformation coefficient of the edge of the fin, the first The distortion coefficient of the fin edge , radiator image The stamping deformation coefficient of the fin edge Both are the same as the first The deformation coefficient of the edge of each fin is positively correlated, and the product of the two can be used to obtain the first Deformation coefficient of the fin edge .

[0087] Step S7: when the deformation coefficient of the fin edge is greater than a preset first threshold, the corresponding fin edge is recorded as a deformed fin edge.

[0088] Specifically, the preset first threshold value may be 0.8. >0.8, the fin edge is recorded as a deformed fin edge, and the corresponding radiator is recorded as a failed product in the fin deformation test item; on the contrary, the deformation coefficient of the fin edge When ≤0.8, the edge of the fin is considered to meet the production standards, and the corresponding radiator is recorded as a qualified product in the fin deformation detection item.

[0089] In the second aspect, this embodiment provides a device for inspecting the quality of heat sinks. Figure 2 , which shows a block diagram of a device for inspecting the quality of a heat sink according to an embodiment of the present invention, the device comprising:

[0090] An image acquisition module 101 is used to acquire a surface image of one side of a fin of a heat sink to be inspected through a camera, and obtain a heat sink image after grayscale processing;

[0091] A stamping deformation coefficient acquisition module 102 is used to acquire the stamping deformation coefficient of each fin edge according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge;

[0092] The distortion coefficient acquisition module 103 is used to acquire the distortion coefficient of each fin edge according to the distance between a fin edge and its nearest fin edge in the heat sink image, the average distance between the fin edge and its nearest fin edge, and the number of pixels on the right and left sides of the fin edge and its fitting straight line;

[0093] The distortion condition acquisition module 104 is used to acquire the distortion condition of each fin edge according to the absolute value of the difference in fitting loss value between a fin edge and the next fin edge in the distortion direction of the fin edge in the heat sink image, and the absolute value of the difference in distance between a fin edge and the next fin edge in the distortion direction and the adjacent fin edge in the distortion direction;

[0094] The distortion coefficient acquisition module 105 is used to acquire the distortion coefficient of each fin edge according to the distortion coefficient and distortion condition of each fin edge;

[0095] The deformation coefficient acquisition module 106 is used to acquire the deformation coefficient of each fin edge according to the twisting deformation coefficient and the stamping deformation coefficient of the fin edge;

[0096] The alarm module 107 records the corresponding fin edge as a deformed fin edge and issues a failure alarm when the deformation coefficient of the fin edge is greater than a preset first threshold.

[0097] Specifically, the first threshold value of the deformation coefficient of the fin edge can be preset to be 0.8. >0.8, the fin edge is recorded as a deformed fin edge, and the corresponding radiator is recorded as a failed product in the fin deformation test item; on the contrary, the deformation coefficient of the fin edge When ≤0.8, the fin edge is considered to meet the production standard, and the corresponding heat sink is recorded as a qualified product in the fin deformation detection item. In addition, when the alarm module performs an unqualified alarm, it can be in the form of an audible and visual alarm, which is not specifically limited in this embodiment.

[0098] The present embodiment provides a method and device for inspecting the quality of heat sink production. First, a camera is used to obtain a surface image of one side of a fin of the heat sink to be inspected, and a heat sink image is obtained after grayscale processing. Then, the stamping deformation coefficient of each fin edge is obtained according to the average difference between the turning degree between a fin edge pixel point and its next pixel point in the heat sink image and the difference between the average grayscale values ​​on both sides of the turning point of the fin edge. The distortion of each fin edge is obtained according to the distance between a fin edge and its nearest fin edge in the heat sink image, the average distance between the fin edge and its nearest fin edge, and the number of pixels on the right and left sides of the fin edge and its fitting straight line. Coefficient; according to the absolute value of the difference in fitting loss values ​​between a fin edge and the next fin edge in the twisting direction of the fin edge in the radiator image, and the absolute value of the difference in distance between a fin edge and the next fin edge in the twisting direction and the adjacent fin edge, obtain the twisting condition of each fin edge; according to the twisting coefficient and twisting condition of each fin edge, obtain the twisting deformation coefficient of each fin edge; finally, according to the twisting deformation coefficient and stamping deformation coefficient of the fin edge, obtain the deformation coefficient of each fin edge; when the deformation coefficient of the fin edge is greater than the preset first threshold, the corresponding fin edge is recorded as a deformed fin edge. The above scheme can solve the problem that the manual quality inspection of the radiator is inefficient in the prior art, and the difficulty of using machine vision to inspect the radiator leads to poor quality inspection results.

[0099] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0100] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for inspecting the quality of heat sink production, characterized in that: The method comprises: Step S1: using a camera to obtain a surface image of one side of the fin of the heat sink to be inspected, and performing grayscale processing to obtain a heat sink image; Step S2: Obtain the stamping deformation coefficient of each fin edge according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge; Step S3: Obtain the distortion coefficient of each fin edge according to the distance between a fin edge and its nearest fin edge in the heat sink image, the average distance between the fin edge and its nearest fin edge, and the number of pixels to the right and left of the fin edge and its fitting line; Step S4: Obtain the distortion of each fin edge according to the absolute value of the difference in fitting loss values ​​between a fin edge and the next fin edge in the fin edge distortion direction in the heat sink image, and the absolute value of the difference between the distances of a fin edge and the next fin edge in the fin edge distortion direction and the adjacent fin edge in the distortion direction; Step S5: obtaining the distortion coefficient of each fin edge according to the distortion coefficient and distortion condition of each fin edge; Step S6: obtaining the deformation coefficient of each fin edge according to the distortion deformation coefficient and the stamping deformation coefficient of the fin edge; Step S7: when the deformation coefficient of the fin edge is greater than a preset first threshold, the corresponding fin edge is recorded as a deformed fin edge; After step S1 and before step S2, the method further includes: filtering out the fin edges in the radiator image from all the edges in the radiator image according to the similarity between the fin edges in the radiator image; The method of selecting the fin edges in the heat sink image from all the edges in the heat sink image according to the similarity between the fin edges in the heat sink image is specifically as follows: according to any edge in the heat sink image and its similar edge, and the distance between any edge in the heat sink image and its similar edge, the fin coefficient of each edge in the heat sink image is calculated, and when the fin coefficient of an edge in the heat sink image is greater than a preset second threshold, the edge is recorded as a fin edge; In step S2, the stamping deformation coefficient of each fin edge is obtained according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge: The angle between any pixel point on the edge of the fin and the line segment formed by its two adjacent edge pixels on the left and right is recorded as the turning degree of the pixel point on the edge of the fin; Compare the turning degree of the pixel point on the edge of the fin with its adjacent pixel points. If the difference between the turning degrees of the pixel point and the pixel points on both sides of it is greater than a third threshold, record it as a turning point. The stamping deformation coefficient of each fin edge is obtained according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge.

2. The method for inspecting the quality of heat sinks according to claim 1, characterized in that: The greater the difference in the turning degree between the fin edge pixel point and the next pixel point in the heat sink image, and the greater the difference in the grayscale value mean on both sides of the turning point of the fin edge, the greater the stamping deformation coefficient of the corresponding fin edge.

3. The method for inspecting the quality of heat sinks according to claim 2, characterized in that: In step S3, when the absolute value of the difference between the distance between a fin edge and its nearest fin edge and the average value of the distance between the fin edge and its nearest fin edge is larger, the distortion coefficient of the fin edge is larger; When the ratio between the number of pixels on the right and left sides of the fin edge and its fitting straight line approaches a constant of 1, the distortion coefficient of the fin edge is smaller.

4. The method for inspecting the quality of heat sinks according to claim 1, characterized in that: In step S4, the smaller the absolute value of the difference in fitting loss values ​​between a fin edge and the next fin edge in the twisting direction of the fin edge in the radiator image, and the smaller the absolute value of the difference in distance between a fin edge and the next fin edge in the twisting direction and the adjacent fin edges in the twisting direction, the more serious the distortion of the fin edge.

5. The method for inspecting good quality of heat sink production according to claim 1, characterized in that: In the step S5, when the distortion coefficient of the fin edge is larger and the distortion is more serious, the distortion deformation coefficient of the fin edge is larger.

6. The method for inspecting good quality of heat sink production according to claim 1, characterized in that: In the step S6, when the twisting deformation coefficient of the fin edge is larger and the stamping deformation coefficient is larger, the deformation coefficient of the fin edge is larger.

7. A radiator production quality inspection device, characterized in that: The device comprises: An image acquisition module is used to acquire a surface image of one side of the fin of the radiator to be inspected through a camera, and obtain a radiator image after grayscale processing; A stamping deformation coefficient acquisition module is used to obtain the stamping deformation coefficient of each fin edge according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge; A distortion coefficient acquisition module is used to obtain the distortion coefficient of each fin edge according to the distance between a fin edge and its nearest fin edge in the heat sink image, the average distance between the fin edge and its nearest fin edge, and the number of pixels on the right and left sides of the fin edge and its fitting straight line; A distortion condition acquisition module is used to acquire the distortion condition of each fin edge according to the absolute value of the difference in fitting loss values ​​between a fin edge and the next fin edge in the distortion direction of the fin edge in the heat sink image, and the absolute value of the difference in distance between a fin edge and the next fin edge in the distortion direction and the adjacent fin edge in the distortion direction; A distortion deformation coefficient acquisition module is used to acquire the distortion deformation coefficient of each fin edge according to the distortion coefficient and distortion condition of each fin edge; A deformation coefficient acquisition module is used to acquire the deformation coefficient of each fin edge according to the twisting deformation coefficient and the stamping deformation coefficient of the fin edge; An alarm module, when the deformation coefficient of the fin edge is greater than a preset first threshold, records the corresponding fin edge as a deformed fin edge and performs a non-conforming alarm; The stamping deformation coefficient acquisition module is also used to filter out the fin edges in the radiator image from all the edges in the radiator image according to the similarity between the fin edges in the radiator image; The method of selecting the fin edges in the heat sink image from all the edges in the heat sink image according to the similarity between the fin edges in the heat sink image is specifically as follows: according to any edge in the heat sink image and its similar edge, and the distance between any edge in the heat sink image and its similar edge, the fin coefficient of each edge in the heat sink image is calculated, and when the fin coefficient of an edge in the heat sink image is greater than a preset second threshold, the edge is recorded as a fin edge; The stamping deformation coefficient acquisition module is used to obtain the stamping deformation coefficient of each fin edge according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge: The angle between any pixel point on the edge of the fin and the line segment formed by its two adjacent edge pixels on the left and right is recorded as the turning degree of the pixel point on the edge of the fin; Compare the turning degree of the pixel point on the edge of the fin with its adjacent pixel points. If the difference between the turning degrees of the pixel point and the pixel points on both sides of it is greater than a third threshold, record it as a turning point. The stamping deformation coefficient of each fin edge is obtained according to the average difference of the turning degree between the fin edge pixel point and the next pixel point in the heat sink image and the difference of the gray value averages on both sides of the turning point of the fin edge.

Images