Defect detection system and method for glue dripping mold

Abstract

The invention discloses a defect detection system and method for a glue dripping mold, and relates to the technical field of glue dripping mold defect detection, and the system comprises a data acquisition module, a reference form transformation curve acquisition module, a form transformation signal generation module, a reference radian coefficient acquisition module, a deformation corner marking module and a corner defect type determination module. The technical problem that complex defects, such as warping, collapse and dislocation defects, of four corner parts of the mold cannot be effectively detected and recognized due to the fact that a current mold quality detection mode lacks comprehensive analysis on the relative position relation of an inner mold and an outer mold is solved. The radian data of the four corners of the die to be detected are collected in real time, the real-time radian coefficient is compared with the corresponding reference radian coefficient, the deformed corners are marked, and the real-time radian distance of the deformed corners is compared with the calculated radian distance, so that the deformed corners can be further subdivided into three types of warping, collapse and dislocation; and automatic classification of defects is realized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of drop glue mold defect detection, and particularly relates to a drop glue mold defect detection system and method. BACKGROUND

[0002] The drop glue mold is widely used in electronic component packaging, process ornament manufacturing and precise pouring process, and the double-layer drop glue mold is composed of an inner mold and an outer mold, the inner mold is used for defining the main forming space of the drop glue, and the outer mold provides external constraint and support. However, due to uneven stress, temperature change, material aging or repeated opening and closing, the mold is prone to deformation, corner warping, collapse or misplacement and other defects in the long-term use process, which will directly lead to poor appearance, size deviation or insufficient structural strength of the formed product, thereby affecting product yield and production efficiency. At present, the quality detection mode of the mold mainly depends on manual visual inspection or uses a projector, a caliper and other single-point measurement tools, which is easily affected by subjective factors and is difficult to realize high-precision and batch automatic detection. There is a lack of comprehensive analysis of the relative position relationship between the inner mold and the outer mold, which leads to the fact that the complex defects of the four corners of the mold, such as warping, collapse and misplacement, cannot be effectively detected and recognized. Therefore, a drop glue mold defect detection system and method are provided. SUMMARY

[0003] The application aims to provide a drop glue mold defect detection system and method, which solves the technical problem that the current quality detection mode of the mold lacks comprehensive analysis of the relative position relationship between the inner mold and the outer mold, leading to the fact that the complex defects of the four corners of the mold, such as warping, collapse and misplacement, cannot be effectively detected and recognized.

[0004] A drop glue mold defect detection system comprises: A reference shape transformation curve acquisition module is configured to analyze the inner mold and outer mold shape data of a plurality of qualified drop glue molds and generate a reference shape transformation curve. A shape transformation signal generation module is configured to obtain a real-time shape transformation curve of a to-be-detected mold according to real-time inner mold and outer mold shape data of the to-be-detected drop glue mold, compare the real-time shape transformation curve with the reference shape transformation curve, and obtain a mold deformation coefficient. When the mold deformation coefficient exceeds a preset threshold Y1, a shape transformation signal is generated, otherwise, no processing is performed. A reference radian coefficient acquisition module is configured to analyze the radian line data of the four corners of a plurality of qualified drop glue molds, obtain the reference radian distance and the reference positioning line slope corresponding to the four corners respectively, and generate the reference radian coefficient corresponding to the four corners respectively based on the reference radian distance and the reference positioning line slope. The deformed corner marking module analyzes the radius data of the four corners of the drop-molding mold in real time when the defect signal is generated, obtains the real-time radius distance, real-time positioning line slope and real-time radius coefficient of the four corners of the drop-molding mold respectively, and compares the real-time radius coefficient of each corner with the corresponding reference radius coefficient to mark the deformed corner. The corner defect type determination module determines the deformed corner as a warping defect, a collapse defect or a misplacement defect according to the relationship between the real-time radius distance of the deformed corner and the corresponding calculated radius distance.

[0005] As a further scheme of the present application, the specific method for generating the reference shape transformation curve is as follows: A1: randomly selecting one of the qualified molds as an analysis mold; A2: taking the distance between the end points of the left and right end wing edges of the inner mold of the analysis mold as the inner mold wing edge extension margin B, taking the end point of the right end wing edge as the starting point, setting a plurality of calibration lines at a fixed calibration distance along the direction of the wing edge extension margin B, obtaining the intersection points between each calibration line and the shape lines of the inner and outer molds of the drop-molding mold respectively, and marking them as inner mold points MAi and outer mold points MBi respectively, calculating the distance between the inner mold points and the outer mold points at each calibration line in the two-dimensional coordinate system and marking it as the inner-outer mold distance Di of the analysis mold at each calibration line, wherein i represents different calibration lines and also serves as the corresponding label of each calibration line, i=1, 2, …, a, and a is the total number of calibration lines; A3: repeating the above steps A1-A2 for all qualified molds, so as to obtain the inner-outer mold distance Dij of each mold at each calibration line, wherein j represents different qualified drop-molding molds; A4: taking the median of the inner-outer mold distances of all qualified molds at the same calibration line as the shape interval, and the shape interval Ji of the drop-molding mold at each calibration line; A5: taking the label i as the horizontal coordinate and the shape interval Ji as the vertical coordinate, drawing and connecting all shape coordinate points Zi(i, Ji), and then obtaining the reference shape transformation curve.

[0006] As a further scheme of the present application, the specific method for obtaining the inner-outer mold distance Di of the analysis mold at each calibration line is as follows: obtaining the sum of the square of the horizontal coordinate difference and the square of the vertical coordinate difference of the inner mold points and the outer mold points of the analysis mold at the same calibration line, performing square root operation on the sum, and taking the obtained value as the inner-outer mold distance of the analysis mold at the calibration line, and then obtaining the inner-outer mold distance Di of the analysis mold at each calibration line.

[0007] As a further scheme of the present application, the specific method for obtaining the mold deformation coefficient is as follows: The inner and outer mold shape data of the drop glue mold to be detected is acquired in real time, and is input into the reference shape transformation curve acquisition module. The inner and outer mold distances at each calibration line are calculated as real-time shape intervals. Taking the label i as the horizontal coordinate and the real-time shape interval as the vertical coordinate, all the shape coordinate points corresponding to the detected drop glue mold are drawn and connected, and then the real-time shape transformation curve of the mold to be detected is obtained. The number of shape coordinate points that do not coincide with the reference curve is counted, and the ratio between the number c and the total number a of calibration lines is taken as the mold deformation coefficient.

[0008] As a further scheme of the present application: the specific way of generating the reference radian coefficients corresponding to the four corners respectively is: A01: randomly select one from the qualified molds as the target mold; The midpoint of the radian line at the four corners of the inner and outer molds of the target mold is marked as the inner and outer radian points, and the inner and outer radian points at the same corner are connected as the radian positioning lines corresponding to the four corners respectively. The reference radian distance and the reference positioning line slope at each corner of the target mold are obtained by analyzing the inner and outer radian point coordinates constituting each radian positioning line. A02: repeat step A01 for all qualified molds, and obtain the reference radian distance and the reference positioning line slope corresponding to the four corners of each qualified drop glue mold respectively. Calculate the average of the reference radian distance and the reference positioning line slope of all qualified molds at the same corner to obtain the calculated radian distance and the calculated positioning line slope corresponding to the four corners respectively. The sum of the product of the calculated radian distance and the calculated positioning line slope corresponding to the same corner and the preset coefficients β1 and β2 is taken as the reference radian coefficient of the corresponding corner, and then the reference radian coefficients corresponding to the four corners respectively are obtained, 1=β1+β2 and β1<β2.

[0009] As a further scheme of the present application: the specific way of obtaining the reference radian distance and the reference positioning line slope at each corner of the target mold is: A two-dimensional coordinate system is established to obtain the inner and outer radian point coordinates constituting each radian positioning line. The same analysis method as obtaining the inner and outer mold distances of the analysis mold at each calibration line is used to obtain the reference radian distance corresponding to the four corners of the target mold. The ratio between the absolute value of the vertical coordinate difference and the absolute value of the horizontal coordinate difference between the inner and outer radian points constituting the same radian positioning line is taken as the reference positioning line slope corresponding to the four corners of the target mold.

[0010] As a further scheme of the present application: the specific way of marking the deformed corner is: The real-time acquired curvature data of the four corners of the epoxy resin mold to be tested is analyzed and imported into the reference curvature coefficient acquisition module to obtain the real-time curvature distance and real-time positioning line slope corresponding to the four corners of the epoxy resin mold to be tested. At the same time, in the same way as obtaining the reference curvature coefficients corresponding to the four corners in step A02, the real-time curvature distance and real-time positioning line slope corresponding to the four corners of the epoxy resin mold to be tested are calculated to obtain the real-time curvature coefficients corresponding to the four corners of the epoxy resin mold to be tested. Corners with real-time curvature coefficients greater than their corresponding reference curvature coefficients are marked as deformed corners, and otherwise, no processing is performed.

[0011] As a further aspect of the present invention, the specific method for determining the deformed corner as a warping defect, collapse defect, or misalignment defect is as follows: Obtain the real-time radian distance of the deformed corner and its corresponding calculated radian distance. Mark the deformed corner with a real-time radian distance greater than its corresponding calculated radian distance as a warping defect corner, mark the deformed corner with a real-time radian distance less than its corresponding calculated radian distance as a collapse defect corner, and mark the deformed corner with a real-time radian distance equal to its corresponding calculated radian distance as a collapse defect corner or a misaligned corner.

[0012] A defect detection method for epoxy resin molds, specifically including the following steps: Step 1: Used to analyze the inner and outer mold morphology data of multiple qualified epoxy resin molds and generate a baseline morphology transformation curve; Step 2: Based on the real-time inner and outer mold morphology data of the epoxy resin mold to be tested, obtain the real-time morphology transformation curve of the mold to be tested, and compare and analyze it with the reference morphology transformation curve to obtain the mold deformation coefficient. When the mold deformation coefficient exceeds the preset threshold Y1, a morphology transformation signal is generated; otherwise, no processing is performed. Step 3: Analyze the arc line data at the four corners of multiple qualified epoxy resin molds to obtain the reference arc distance and reference positioning line slope corresponding to each of the four corners. Generate the reference arc coefficient corresponding to each of the four corners based on the reference arc distance and reference positioning line slope. Step 4: When a defect signal is generated, the real-time acquired curvature data of the four corners of the die to be tested is analyzed to obtain the real-time curvature distance, real-time positioning line slope, and real-time curvature coefficient of the four corners of the die to be tested. The real-time curvature coefficient of each corner is compared with the corresponding reference curvature coefficient, and the deformed corners are marked. Step 5: Based on the relationship between the real-time arc distance of the deformed corner and the corresponding calculated arc distance, determine the deformed corner as a warping defect, a collapse defect, or a misalignment defect.

[0013] Compared with the prior art, the beneficial effects of the present invention are: (1) In this invention, by comparing the overlap between the real-time curve and the reference curve, the proportion of non-overlapping points to the total number of points is used as the mold deformation coefficient. When the deformation coefficient exceeds the preset threshold Y1, the system outputs a shape transformation signal, which realizes the real-time and automated judgment of the overall structural deviation of the mold. It can quickly determine whether the mold as a whole has deformed, efficiently screen out obviously unqualified products, and serve as the first checkpoint to trigger subsequent fine analysis. (2) In this invention, by selecting inner and outer arc caliber points at the four corners of a qualified mold to construct arc positioning lines, calculating the arc distance and the slope rate of the positioning lines, and by statistically calculating multiple qualified molds and combining them with preset weight coefficients β1 and β2, the reference arc coefficients of the four corners can be obtained, which can reflect the geometric health status of the key stress points of the mold and provide a reference for corner defect detection. (3) In this invention, when a defect signal is generated, the arc data of the four corners of the mold to be tested is collected in real time, and the real-time arc coefficient is compared with the corresponding reference arc coefficient. The deformed corner is marked, and the accurate positioning and detection of the four key corners of the mold is realized. The abnormal position can be found quickly, avoiding the problem that the traditional method can only find the overall deviation but it is difficult to locate the specific defect. (4) In this invention, by comparing the real-time arc distance of the deformed corner with the calculated arc distance, if the real-time value is greater than the reference value, it is determined to be a warping defect; if the real-time value is less than the reference value, it is determined to be a collapse defect; if the real-time value is equal to the reference value but the direction is offset, it is determined to be a misalignment defect; the deformed corner can be further subdivided into three types: warping, collapse and misalignment, so as to realize the automatic classification of defects. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the system framework structure of the present invention; Figure 2 This is a schematic diagram of the method framework structure of the present invention; Figure 3 This is a schematic diagram of the calibration lines, inner mold points, and outer mold points of the present invention; Figure 4 This is a schematic diagram of the inner arc caliper, outer arc caliper, and arc positioning line structure of the present invention. Detailed Implementation

[0015] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example 1: Please refer to Figure 1 ,Figure 3 and Figure 4 This application provides a defect detection system for epoxy resin molds, including: The data acquisition module acquires the inner and outer mold morphology data of multiple qualified epoxy resin molds. The inner and outer mold morphology data of epoxy resin molds refer to the edge morphology lines corresponding to the inner and outer molds of the epoxy resin mold. The structured light scanner, such as a laser profilometer, can be used to acquire the three-dimensional point cloud of the cross section, or the contour tracking algorithm can be used to extract the boundary contour of the inner and outer molds. The above are existing technologies, so they will not be described in detail here. By utilizing existing high-precision sensing technologies or image processing algorithms, a static or moving double-layer epoxy resin mold is scanned or photographed to obtain three-dimensional point cloud data or two-dimensional images of its cross-section. From these images, the edge contour lines representing the shapes of the inner and outer molds are precisely extracted. These contour lines serve as the data source for all subsequent analyses, enabling non-contact and automated data acquisition. This replaces manual measurement and lays the foundation for fully automated inspection lines. By employing advanced technologies such as structured light scanning, the acquired data is highly accurate and rich in detail, providing reliable data support for subsequent precision analysis. This method can quickly and non-contactly acquire complete morphological information of the mold, improving inspection efficiency.

[0017] The baseline morphology transformation curve acquisition module analyzes the inner and outer mold morphology data of multiple qualified epoxy resin molds to obtain the baseline morphology transformation curve corresponding to the epoxy resin mold. The specific method is as follows: Obtain cross-sectional images of each qualified epoxy resin mold, and extract the cross-sectional morphology images of each qualified epoxy resin mold. The cross-sectional morphology images include the morphology lines corresponding to the inner mold and the outer mold in the epoxy resin mold, respectively. A1: Randomly select one of several qualified epoxy resin molds as the analysis mold; A2: Obtain the distance between the endpoints of the left and right wings of the inner mold of the analysis mold, and use it as the extension distance B of the inner mold wing. Taking the endpoint of the right wing of the inner mold of the epoxy resin mold as the starting point, set multiple calibration lines Li along the direction of the extension distance B of the wing at a fixed calibration distance. At the same time, obtain the intersection points between each calibration line Li and the shape lines of the inner and outer molds of the epoxy resin mold, and mark them as the inner mold point MAi and the outer mold point MBi respectively (e.g., Figure 3 The specific value of the calibration distance is determined by relevant staff based on actual needs. Here, i represents different calibration lines and serves as the label corresponding to each calibration line. i = 1, 2, ..., a, where a is the total number of calibration lines. a is a positive integer and a1 is the integer value of the ratio between the extended edge distance B along the wing edge and the calibration distance. The extension distance B of the inner mold wing edge can be obtained by manual measurement or distance detector, which is existing technology and will not be described in detail here; A two-dimensional coordinate system is set up within the cross-sectional image of the mold for analysis. The coordinates of the inner mold point MAi and the outer mold point MBi corresponding to each calibration line Li are obtained. Based on the coordinates of the inner mold point MAi and the outer mold point MBi located at the same calibration line, the distance between the inner mold point and the outer mold point corresponding to each calibration line is calculated using the Euclidean distance calculation method. The specific method for obtaining the distance between the inner and outer molds corresponding to each calibration line is as follows: The sum of the squares of the differences in the abscissa and ordinate of the inner and outer mold points located at the same calibration line is obtained, and the square root of this sum is calculated. The resulting value is used as the distance between the inner and outer molds of the analysis mold located at the calibration line, thereby obtaining the distance Di between the inner and outer molds of the analysis mold located at each calibration line. A3: Repeat steps A1-A2 to obtain the inner and outer mold distances Dij corresponding to each qualified epoxy resin mold at each calibration line, where j represents different qualified epoxy resin molds; A4: Obtain the median of the distance between the inner and outer molds of each qualified epoxy resin mold at the same calibration line, that is, the average of the maximum and minimum values ​​of the distance between the inner and outer molds at the same calibration line; and use it as the morphological spacing Ji of the epoxy resin mold at each calibration line. A5: Based on the shape spacing Ji corresponding to the inner and outer mold shape lines at each calibration line, obtain the reference shape transformation curve XA corresponding to the epoxy mold. The specific method is as follows: Using the label i corresponding to each calibration line as the abscissa and the morphological spacing Ji corresponding to each calibration line of the dispensing mold as the ordinate, the morphological coordinate points Zi (i, Ji) corresponding to each calibration line are plotted in the two-dimensional coordinate system. The morphological coordinate points Zi are connected in the order from front to back to obtain the reference morphological transformation curve corresponding to the dispensing mold. By setting a series of fixed-interval calibration lines along the extended edge of the inner mold, the distance between the inner and outer molds on these calibration lines is calculated for each qualified mold. Then, by calculating the median distance between the inner and outer molds at the same calibration line for all qualified molds, individual differences are eliminated to obtain the morphological spacing. Finally, with the label as the abscissa and the morphological spacing as the ordinate, all points are plotted and connected to form a reference morphological transformation curve representing the standard morphology of a qualified mold. This curve is used to analyze the morphological data of the inner and outer molds of multiple qualified molds. The reference morphological transformation curve reflects the change in the distance between the inner and outer molds, enabling subsequent inspections to be based on a standardized qualified model, thus ensuring the objectivity and consistency of the inspection results.

[0018] The morphology transformation signal generation module acquires the morphology data of the inner and outer molds of the epoxy resin mold to be tested in real time, obtains the real-time morphology transformation curve of the mold, compares and analyzes the real-time morphology transformation curve with the reference morphology transformation curve, and generates a morphology transformation signal based on the analysis results. Specifically: Data from the epoxy resin mold to be tested is acquired in real time and input into the baseline morphological transformation curve acquisition module. Using the same method as obtaining the inner and outer mold distances of the analysis mold at each calibration line in step A2, these distances are acquired and used as the real-time morphological spacing of the epoxy resin mold at each calibration line. Then, using the same method as obtaining the baseline morphological transformation curve corresponding to the epoxy resin mold in step A5, the label i corresponding to each calibration line is used as the abscissa, and the real-time morphological spacing of the epoxy resin mold at each calibration line is used as the ordinate. Obtain the morphological coordinate points corresponding to each calibration line of the epoxy resin mold to be tested. At the same time, draw the real-time morphological transformation curve of the epoxy resin mold to be tested in the two-dimensional coordinate system where the reference morphological transformation curve is located, based on the morphological coordinate points of each morphological coordinate point of the epoxy resin mold to be tested. Obtain the number c of morphological coordinate points that do not coincide with the reference morphological transformation curve in the real-time morphological transformation curve. The ratio between the number c and the total number a of the calibration lines is used as the mold deformation coefficient. When the mold deformation coefficient is greater than the preset threshold Y1, a morphological transformation signal is generated. Otherwise, no processing is performed. The preset threshold Y1 is greater than 1 / 2 and less than 3 / 5. The mold to be inspected undergoes the same calibration and calculation steps as the reference curve, and a real-time shape transformation curve is plotted. The overlap between the real-time curve and the reference curve is compared, and the proportion of non-overlapping points to the total number of points is used as the mold deformation coefficient. When the deformation coefficient exceeds the preset threshold Y1, the system outputs a shape transformation signal. This achieves real-time and automated judgment of the overall structural deviation of the mold, avoiding manual point-by-point comparison and improving inspection efficiency. At the same time, the mold deformation coefficient quantifies the degree of deviation, facilitating threshold control and batch inspection. ， It can quickly determine whether a mold has undergone overall deformation and efficiently screen out obviously defective products, serving as the first step to trigger subsequent detailed analysis.

[0019] Example 2: As Example 2 of the present invention, in specific implementation, the technical solution of this example differs from that of Example 1 only in that this example also includes a reference radian coefficient acquisition module and a deformation corner marking module; The reference arc coefficient acquisition module acquires the arc line data corresponding to the four corners of multiple qualified dispensing molds, analyzes the data to obtain the reference arc distance and reference positioning line slope corresponding to the four corners of the dispensing mold, and obtains the reference arc coefficient corresponding to the four corners of the dispensing mold based on the analysis of the reference arc distance and reference positioning line slope corresponding to the four corners of the dispensing mold. The specific method for obtaining the reference arc distance and the slope of the reference positioning line at the four corners of the epoxy mold is as follows: A01: Randomly select one of several qualified epoxy resin molds as the target mold; Mark the midpoints of the arc lines at the four corners of the inner and outer molds of the target mold as inner and outer arc markers. Connect the inner and outer arc markers located at the same corner to obtain the arc positioning lines corresponding to the four corners of the inner and outer molds of the target mold (e.g., ...). Figure 4 ); A two-dimensional coordinate system is established to obtain the coordinates of the inner and outer arc points that make up each arc positioning line. Based on the coordinates of the inner and outer arc points that make up each arc positioning line, the same method is used to obtain the distance between the inner and outer molds corresponding to the analysis mold at each calibration line to obtain the reference arc distance Me1 corresponding to the four corners of the target mold. At the same time, the ratio between the absolute value of the difference between the vertical coordinate and the absolute value of the difference between the horizontal coordinate of the inner and outer arc points that make up the same arc positioning line is used as the slope Ke1 of the reference positioning line corresponding to the four corners of the target mold, where e represents different corners and is also the corner number corresponding to different corners, e=1, 2, 3, 4; A02: Repeat step A01 to obtain the reference radian distance Mej and reference positioning line slope Kej corresponding to the four corners of each qualified epoxy resin mold. Take the average value of each reference radian distance and reference positioning line slope corresponding to the same corner of each qualified epoxy resin mold as the calculated radian distance JA1 and calculated positioning line slope JB1 at the corresponding corner. Take the sum of the products of the calculated radian distance JA and the calculated positioning line slope JB with the preset coefficients β1 and β2 respectively as the reference radian coefficient H1 of the epoxy resin mold at the corresponding corner. Use the same method to analyze the reference radian distance and reference positioning line slope corresponding to each of the remaining corners, and then obtain the calculated radian distance JAe and calculated positioning line slope JBe corresponding to the four corners of the epoxy resin mold. At the same time, obtain the reference radian coefficient He corresponding to the four corners of the epoxy resin mold. The specific values ​​of the preset coefficients β1 and β2 are determined by relevant personnel according to actual needs. Also, 1 = β1 + β2, and β1 < β2. At the four corners of a qualified mold, inner and outer radii are selected to construct radii positioning lines, and the radii distance and the slope rate of the positioning lines are calculated. By statistically averaging multiple qualified molds and combining them with preset weighting coefficients β1 and β2, the baseline radii coefficients of the four corners are obtained, which can reflect the geometric health status of the key stress points of the mold and provide a reference for corner defect detection.

[0020] The deformation corner marking module, when a defect signal is generated, acquires the real-time arc data corresponding to the four corners of the epoxy resin mold to be inspected, analyzes it to obtain the real-time arc coefficients corresponding to the four corners of the mold, and compares the real-time arc coefficients corresponding to the four corners with their corresponding reference arc coefficients to mark the deformed corners. The specific method is as follows: The arc line data corresponding to the four corners of the epoxy resin mold to be tested are imported into the reference arc coefficient acquisition module. The data is analyzed in the same way as the reference arc distance and reference positioning line slope corresponding to the four corners of the target mold obtained in step A01 to obtain the real-time arc distance and real-time positioning line slope corresponding to the four corners of the epoxy resin mold to be tested. At the same time, the real-time arc distance and real-time positioning line slope corresponding to the four corners of the epoxy resin mold to be tested are calculated in the same way as the reference arc coefficient corresponding to the four corners of the epoxy resin mold obtained in step A02 to obtain the real-time arc coefficient corresponding to the four corners of the epoxy resin mold to be tested. The real-time arc coefficient corresponding to the four corners is compared and analyzed with its corresponding reference arc coefficient. Corners with a real-time arc coefficient greater than their corresponding reference arc coefficient are marked as deformed corners, and otherwise, no processing is performed. When a defect signal is generated, the radii of the four corners of the mold to be inspected are collected in real time. The real-time radii coefficient is calculated and compared with the corresponding reference radii coefficient. When the real-time coefficient is greater than the reference value, the corner is marked as a deformed corner. This achieves precise positioning and detection of the four key corners of the mold, and can quickly find abnormal locations, avoiding the problem that traditional methods can only detect overall deviations but cannot locate specific defective parts.

[0021] Example 3: As Example 3 of the present invention, in specific implementation, compared with Example 1 and Example 2, the technical solution of this example differs from that of Example 1 only in that this example also includes a corner defect type determination module; The corner defect type determination module obtains the real-time radian distance corresponding to the deformed corner from the deformed corner marking module, and at the same time obtains the calculated radian distance corresponding to the deformed corner from the calculated radian distance JAe corresponding to the four corners of the dripping mold. Deformed corners with real-time radian distance greater than their corresponding calculated radian distance are marked as warping defect corners, deformed corners with real-time radian distance less than their corresponding calculated radian distance are marked as collapse defect corners, and deformed corners with real-time radian distance equal to their corresponding calculated radian distance are marked as collapse defect corners and misaligned corners. By comparing the real-time radian distance of the deformed corner with the calculated radian distance, if the real-time value is greater than the reference value, it is determined to be a warping defect; if the real-time value is less than the reference value, it is determined to be a collapse defect; if the real-time value is equal to the reference value but the direction is off, it is determined to be a misalignment defect. The deformed corner can be further subdivided into three types: warping, collapse, and misalignment, realizing automatic defect classification. This not only determines whether a defect exists, but also provides information on the nature of the defect, providing a targeted basis for subsequent maintenance and mold life management.

[0022] Example 4: Please refer to Figure 2 As shown, this embodiment also provides a defect detection method for epoxy resin molds. This method is implemented using the aforementioned defect detection system for epoxy resin molds and specifically includes the following steps; Step 1: Used to analyze the inner and outer mold morphology data of multiple qualified epoxy resin molds and generate a baseline morphology transformation curve; Step 2: Based on the real-time inner and outer mold morphology data of the epoxy resin mold to be tested, obtain the real-time morphology transformation curve of the mold to be tested, and compare and analyze it with the reference morphology transformation curve to obtain the mold deformation coefficient. When the mold deformation coefficient exceeds the preset threshold Y1, a morphology transformation signal is generated; otherwise, no processing is performed. Step 3: Analyze the arc line data at the four corners of multiple qualified epoxy resin molds to obtain the reference arc distance and reference positioning line slope corresponding to each of the four corners. Generate the reference arc coefficient corresponding to each of the four corners based on the reference arc distance and reference positioning line slope. At the four corners of a qualified mold, inner and outer arc markers are selected respectively to construct arc positioning lines. The arc distance and the slope rate of the positioning lines are calculated. By statistical averaging of multiple qualified molds and combining the preset weight coefficients β1 and β2, the reference arc system of the four corners is obtained.

[0023] Step 4: When a defect signal is generated, the real-time acquired curvature data of the four corners of the die to be tested is analyzed to obtain the real-time curvature distance, real-time positioning line slope, and real-time curvature coefficient of the four corners of the die to be tested. The real-time curvature coefficient of each corner is compared with the corresponding reference curvature coefficient, and the deformed corners are marked. Step 5: Based on the relationship between the real-time radian distance of the deformed corner and the corresponding calculated radian distance, determine the deformed corner as a warping defect, a collapse defect, or a misalignment defect; By comparing the real-time radian distance of the deformed corner with the calculated radian distance, if the real-time value is greater than the reference value, it is determined to be a warping defect; if the real-time value is less than the reference value, it is determined to be a collapse defect; if the real-time value is equal to the reference value but the direction is off, it is determined to be a misalignment defect. The deformed corner can be further subdivided into three types: warping, collapse, and misalignment, so as to realize the automatic classification of defects.

[0024] Example 5: As Example 5 of the present invention, in specific implementation, compared with Example 1, Example 2, Example 3 and Example 4, the technical solution of this example is to combine the solutions of Example 1, Example 2, Example 3 and Example 4.

[0025] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0026] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A defect detection system for epoxy resin molds, characterized in that, include: The baseline morphology transformation curve acquisition module is used to analyze the inner and outer mold morphology data of multiple qualified epoxy resin molds and generate baseline morphology transformation curves. The shape transformation signal generation module obtains the real-time shape transformation curve of the mold to be tested based on the real-time inner and outer mold shape data of the dripping mold to be tested, and compares and analyzes it with the reference shape transformation curve to obtain the mold deformation coefficient. When the mold deformation coefficient exceeds the preset threshold Y1, a shape transformation signal is generated; otherwise, no processing is performed. The reference arc coefficient acquisition module is used to analyze the arc line data at the four corners of multiple qualified dispensing molds, obtain the reference arc distance and reference positioning line slope corresponding to the four corners respectively, and generate the reference arc coefficient corresponding to the four corners based on the reference arc distance and reference positioning line slope. The deformation corner marking module analyzes the real-time acquired curvature data of the four corners of the die to be tested when a defect signal is generated. It obtains the real-time curvature distance, real-time positioning line slope, and real-time curvature coefficient of the four corners of the die to be tested. It then compares the real-time curvature coefficient of each corner with the corresponding reference curvature coefficient to mark the deformation corner. The corner defect type determination module determines the deformed corner as a warping defect, collapse defect, or misalignment defect based on the relationship between the real-time arc distance of the deformed corner and the corresponding calculated arc distance.

2. The defect detection system for a dispensing mold according to claim 1, characterized in that, The specific method for generating the baseline morphological transformation curve is as follows: A1: Randomly select one of the multiple qualified molds as the analysis mold; A2: The distance between the left and right wing endpoints of the inner mold is taken as the inner mold wing extension distance B. Taking the right wing endpoint as the starting point, multiple calibration lines are set at fixed calibration distances along the wing extension distance B. The intersection points between each calibration line and the inner and outer mold shape lines of the epoxy resin mold are obtained, and they are marked as inner mold point MAi and outer mold point MBi, respectively. In the two-dimensional coordinate system, the distance between the inner mold point and the outer mold point at each calibration line is calculated and marked as the inner and outer mold distance Di of the analysis mold at each calibration line, where i represents different calibration lines and is also used as the corresponding label for each calibration line, i=1, 2, ..., a, where a is the total number of calibration lines; A3: Repeat steps A1-A2 above for all qualified molds to obtain the inner and outer mold distances Dij corresponding to each mold at each calibration line, where j refers to different qualified epoxy resin molds; A4: The median distance between the inner and outer molds of all qualified molds at the same calibration line is taken as the shape spacing. The shape spacing Ji of the dripping mold at each calibration line is respectively. A5: Using label i as the abscissa and morphological spacing Ji as the ordinate, draw and connect all morphological coordinate points Zi(i, Ji) to obtain the baseline morphological transformation curve.

3. The defect detection system for a dispensing mold according to claim 2, characterized in that, The specific method for obtaining the distance Di between the inner and outer molds at each calibration line is as follows: The sum of the squares of the differences in the abscissa and ordinate of the inner and outer mold points located at the same calibration line is obtained, and the square root of this sum is calculated. The resulting value is used as the distance between the inner and outer molds of the analysis mold at that calibration line, thereby obtaining the distance Di between the inner and outer molds of the analysis mold at each calibration line.

4. The defect detection system for a dispensing mold according to claim 2, characterized in that, The specific method for obtaining the mold deformation coefficient is as follows: The morphological data of the inner and outer molds of the epoxy resin mold to be tested are acquired in real time and input into the reference morphological transformation curve acquisition module. The distance between the inner and outer molds at each calibration line is calculated as the real-time morphological spacing. With the label i as the abscissa and the real-time morphological spacing as the ordinate, all morphological coordinate points corresponding to the epoxy resin mold to be tested are plotted and connected to obtain the real-time morphological transformation curve of the mold to be tested. The number c of morphological coordinate points that do not coincide with the reference curve of the real-time morphological transformation curve is counted. The ratio between the number c and the total number a of the calibration lines is used as the mold deformation coefficient.

5. The defect detection system for a dispensing mold according to claim 4, characterized in that, The specific method for generating the reference radian coefficients corresponding to the four corners is as follows: A01: Randomly select one of the qualified molds as the target mold; Mark the midpoints of the arc lines at the four corners of the inner and outer molds of the target mold as inner and outer arc markers respectively. Connect the inner and outer arc markers located at the same corner to form the arc positioning lines corresponding to the four corners respectively. Analyze the coordinates of the inner and outer arc markers that make up each arc positioning line to obtain the reference arc distance and reference positioning line slope at each corner of the target mold. A02: Repeat step A01 for all qualified molds to obtain the reference radian distance and reference positioning line slope corresponding to the four corners of each qualified epoxy resin mold. Calculate the average of the reference radian distance and reference positioning line slope of all qualified molds at the same corner to obtain the calculated radian distance and calculated positioning line slope corresponding to the four corners. The sum of the products of the calculated radian distance and calculated positioning line slope corresponding to the same corner and the preset coefficients β1 and β2 is used as the reference radian coefficient at the corresponding corner. Thus, the reference radian coefficients corresponding to the four corners are obtained, where 1 = β1 + β2 and β1 < β2.

6. The defect detection system for a dispensing mold according to claim 5, characterized in that, The specific method for obtaining the reference radian distance and reference positioning line slope at each corner of the target mold is as follows: A two-dimensional coordinate system is established to obtain the coordinates of the inner and outer arc points that make up each arc positioning line. Based on the coordinates of the inner and outer arc points that make up each arc positioning line, the same analysis method as obtaining the distance between the inner and outer molds of the analysis mold at each calibration line is adopted to obtain the reference arc distances corresponding to the four corners of the target mold. At the same time, the ratio between the absolute value of the difference in the vertical coordinate and the absolute value of the difference in the horizontal coordinate between the inner and outer arc points that make up the same arc positioning line is used as the slope of the reference positioning line corresponding to the four corners of the target mold.

7. A defect detection system for a dispensing mold according to claim 5, characterized in that, The specific method for marking deformed corners is as follows: The real-time acquired curvature data of the four corners of the epoxy resin mold to be tested is analyzed and imported into the reference curvature coefficient acquisition module to obtain the real-time curvature distance and real-time positioning line slope corresponding to the four corners of the epoxy resin mold to be tested. At the same time, in the same way as obtaining the reference curvature coefficients corresponding to the four corners in step A02, the real-time curvature distance and real-time positioning line slope corresponding to the four corners of the epoxy resin mold to be tested are calculated to obtain the real-time curvature coefficients corresponding to the four corners of the epoxy resin mold to be tested. Corners with real-time curvature coefficients greater than their corresponding reference curvature coefficients are marked as deformed corners, and otherwise, no processing is performed.

8. The defect detection system for a dispensing mold according to claim 7, characterized in that, The specific method for determining whether a deformed corner is a warping defect, a collapse defect, or a misalignment defect is as follows: Obtain the real-time radian distance of the deformed corner and its corresponding calculated radian distance. Mark the deformed corner with a real-time radian distance greater than its corresponding calculated radian distance as a warping defect corner, mark the deformed corner with a real-time radian distance less than its corresponding calculated radian distance as a collapse defect corner, and mark the deformed corner with a real-time radian distance equal to its corresponding calculated radian distance as a collapse defect corner or a misaligned corner.

9. A method for defect detection in epoxy resin molds, characterized in that, This method is implemented using a defect detection system for epoxy resin molds as described in any one of claims 1-8, and specifically includes the following steps: Step 1: Used to analyze the inner and outer mold morphology data of multiple qualified epoxy resin molds and generate a baseline morphology transformation curve; Step 2: Based on the real-time inner and outer mold morphology data of the epoxy resin mold to be tested, obtain the real-time morphology transformation curve of the mold to be tested, and compare and analyze it with the reference morphology transformation curve to obtain the mold deformation coefficient. When the mold deformation coefficient exceeds the preset threshold Y1, a morphology transformation signal is generated; otherwise, no processing is performed. Step 3: Analyze the arc line data at the four corners of multiple qualified epoxy resin molds to obtain the reference arc distance and reference positioning line slope corresponding to each of the four corners. Generate the reference arc coefficient corresponding to each of the four corners based on the reference arc distance and reference positioning line slope. Step 4: When a defect signal is generated, the real-time acquired curvature data of the four corners of the die to be tested is analyzed to obtain the real-time curvature distance, real-time positioning line slope, and real-time curvature coefficient of the four corners of the die to be tested. The real-time curvature coefficient of each corner is compared with the corresponding reference curvature coefficient, and the deformed corners are marked. Step 5: Based on the relationship between the real-time arc distance of the deformed corner and the corresponding calculated arc distance, determine the deformed corner as a warping defect, a collapse defect, or a misalignment defect.

Images