CN115266762A - Tin ball defect detection device and method, electronic device and storage medium - Google Patents

Abstract

The application relates to the technical field of solder ball detection, and discloses a solder ball defect detection device, a solder ball defect detection method, electronic equipment and a storage medium. The tin ball defect detection device comprises a light source assembly, an image acquisition assembly and a processor, wherein the light source assembly comprises a first light source and a second light source, the first light source is used for generating first color light, the second light source is used for generating second color light, the colors of the first color light and the second color light are different, and the first color light and the second color light are used for irradiating a to-be-detected tin soldering piece; the image acquisition assembly is used for acquiring an image to be detected when the soldering tin piece to be detected is irradiated by the first color light and the second color light; the processor is connected with the image acquisition assembly and used for receiving the image to be detected and carrying out image recognition analysis on the image to be detected so as to carry out defect detection on the soldering tin piece to be detected. By the method, the accuracy and the detection efficiency of detecting the defects of the solder balls can be improved.

Description

Solder ball defect detection device, method, electronic equipment and storage medium

technical field

The present application relates to the technical field of solder ball detection, in particular to a solder ball defect detection device, method, electronic equipment and storage medium.

Background technique

In the production of electronic products, the defect detection link is very important to ensure the reliability of product quality. Among them, the printed circuit board (PCB) is the basis for the operation of electronic products, and the solder balls on it play the role of connection and conduction, so it is required that it should not have obvious shape defects so as not to affect the system performance.

A manual identification method is adopted in the related art, but this method is inefficient.

Contents of the invention

The technical problem mainly solved by this application is to provide a solder ball defect detection device, method, electronic equipment and storage medium, which can improve the accuracy and detection efficiency of solder ball defect detection.

In order to solve the above problems, a technical solution adopted by this application is to provide a solder ball defect detection device, the solder ball defect detection device includes: a light source assembly, including a first light source and a second light source, the first light source is used to generate the first One color light, the second light source is used to generate the second color light, the color of the first color light and the second color light are different, the first color light and the second color light are used to irradiate the solder piece to be detected; the image acquisition component uses To collect the image to be inspected when the solder piece to be inspected is irradiated by the first color light and the second color light; the processor is connected to the image acquisition component for receiving the image to be inspected, and performing image recognition and analysis on the image to be inspected, to treat Detect solder parts for defect detection.

Wherein, the light source assembly is provided with an optical channel, and the image acquisition assembly is arranged on the side of the light source assembly away from the solder piece to be inspected. The light generated by the light source assembly is reflected by the solder object to be inspected and then enters the image acquisition assembly through the optical channel.

Wherein, the first light source is provided with a first through hole, the second light source is provided with a second through hole, and the first light source and the second light source are stacked so that the first through hole and the second through hole are connected to form an optical channel; The diameter of the first through hole is smaller than the diameter of the second through hole.

Wherein, the first light source is provided with a first through hole, the second light source is provided with a second through hole, and the second light source is arranged in the first through hole so that the second through hole forms a light channel.

Wherein, the first light source includes a plurality of first sub-light sources, and the plurality of first sub-light sources are arranged in a ring to form a first through hole in the plurality of first sub-light sources; the second light source includes a plurality of second sub-light sources, and a plurality of The second sub-light sources are arranged in a ring shape to form second through holes in the plurality of second sub-light sources.

Wherein, the solder ball defect detection device further includes a support frame, and the light source component and the image acquisition component are arranged on the support frame.

Wherein, the support frame includes a first support member and a second support member; the light source assembly is arranged on the first support member, and the image acquisition assembly is arranged on the second support member; side.

Wherein, the first color light is monochromatic light, and the second color light is polychromatic light or monochromatic light.

Wherein, the first color light is red light, and the second color light is white light.

In order to solve the above problems, another technical solution adopted by this application is to provide a solder ball defect detection method, which includes: obtaining an image to be detected; using a defect detection model to detect the image to be detected, and obtaining the corresponding defect in the image to be detected The detection information, wherein, the defect detection model is obtained by training the training image, wherein the image to be detected and the training image are obtained by using the image acquisition component in the solder ball defect detection device provided by the above technical solution when the solder piece to be detected is divided into two different Acquired when illuminated by colored light.

In order to solve the above problems, another technical solution adopted by the present application is to provide an electronic device, the electronic device includes a controller and a memory connected to the controller; wherein, the memory is used to store program data, and the controller is used to execute the program data, To realize the method provided by the above technical solution.

In order to solve the above problems, another technical solution adopted by this application is to provide a computer-readable storage medium, which is used to store program data. When the program data is executed by the processor, it is used to realize the above-mentioned The method provided by the technical solution.

The beneficial effects of the present application are: different from the situation of the prior art, a solder ball defect detection device of the present application, the solder ball defect detection device includes: a light source assembly, including a first light source and a second light source, the first light source is used For generating the first color light, the second light source is used for generating the second color light, the colors of the first color light and the second color light are different, and the first color light and the second color light are used to irradiate the solder piece to be inspected; image acquisition The component is used to collect the image to be detected when the solder piece to be detected is irradiated by the first color light and the second color light; the processor is connected to the image acquisition component and is used to receive the image to be detected and perform image recognition and analysis on the image to be detected , so as to detect the defects of the solder parts to be tested. Through the above method, two different colors of light are used to irradiate the solder parts to be inspected, so that the images collected by the hemispherical solder balls on the solder parts to be inspected have obvious color differences, which can make the solder balls with defects more prominent. Further, the accuracy of the detection of the solder ball defect is improved, and the detection efficiency of the detection of the solder ball defect is improved.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings that need to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those skilled in the art, other drawings can also be obtained based on these drawings without creative effort. in:

FIG. 1 is a schematic structural view of an embodiment of a solder ball defect detection device provided by the present application;

Fig. 2 is a schematic structural view of an embodiment of the first light source provided by the present application;

Fig. 3 is a schematic structural diagram of an embodiment of the first light source and the second light source provided by the present application;

Fig. 4 is a schematic structural diagram of an embodiment of a light source assembly and an image acquisition assembly provided by the present application;

Fig. 5 is a schematic structural diagram of another embodiment of a light source assembly and an image acquisition assembly provided by the present application;

Fig. 6 is a schematic structural diagram of another embodiment of a light source assembly and an image acquisition assembly provided by the present application;

FIG. 7 is a schematic structural view of another embodiment of the solder ball defect detection device provided by the present application;

Fig. 8 is a schematic structural view of an embodiment of a light source assembly, an image acquisition assembly and a support provided by the present application;

Fig. 9 is a schematic structural diagram of another embodiment of the light source assembly provided by the present application;

Fig. 10 is a schematic diagram of the image to be detected provided by the present application after defect detection;

FIG. 11 is a schematic flow chart of an embodiment of a solder ball defect detection method provided by the present application;

FIG. 12 is a schematic structural diagram of an embodiment of an electronic device provided by the present application;

Fig. 13 is a schematic structural diagram of an embodiment of a computer-readable storage medium provided by the present application.

Detailed ways

The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the specific embodiments described here are only used to explain the present application, but not to limit the present application. In addition, it should be noted that, for the convenience of description, only some structures related to the present application are shown in the drawings but not all structures. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.

The terms "first", "second", etc. in this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "include" and "have", as well as any variations thereof, are intended to cover a non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but optionally also includes unlisted steps or units, or optionally further includes For other steps or units inherent in these processes, methods, products or apparatuses.

Reference herein to an "embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application. The occurrences of this phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is understood explicitly and implicitly by those skilled in the art that the embodiments described herein can be combined with other embodiments.

Solder parts can be composed of circuit boards and electronic components. Wherein, the electronic components are fixed on the circuit board by soldering. For example, the solder part can be a chip. However, there are solder ball defects during soldering, and the solder ball defects will cause subsequent solder parts to fail and affect performance. Therefore, it is necessary to detect solder ball defects on solder parts.

In the related art, a manual method is used to detect solder ball defects. Based on this, the following embodiments are provided to detect solder ball defects, so as to improve detection efficiency.

Referring to FIG. 1 , FIG. 1 is a schematic structural diagram of an embodiment of a solder ball defect detection device provided in the present application. The solder ball defect detection device 100 includes a light source assembly (not shown), an image acquisition assembly 20 and a processor 30 .

Wherein, the light source assembly includes a first light source and a second light source, and the first light source is used to generate light of the first color. The second light source is used to generate light of the second color.

Wherein, the first color light and the second color light have different colors, and the first color light and the second color light are used to irradiate the solder piece to be inspected. For example, the first color light is red, and the second color light is white. In another example, the first color light is blue, and the second color light is white. Specifically, the color light generated by the first light source and the second light source can be correspondingly set according to the color of the solder piece to be detected. For example, if the surface of the substrate of the solder piece is black, the first color light is red, and the second color light is white.

In some embodiments, the first light source and the second light source can generate multiple colors of light, so that the corresponding color light can be selected as required. In an application scenario, referring to FIG. 2 , the first light source 11 includes a plurality of lamp beads 111 . Part of the lamp beads 111 can emit light of the same color, and the rest of the lamp beads 111 can emit light of another same color. The corresponding lamp bead 111 can be selected to emit light of the first color according to the color of the substrate in the solder piece to be detected. It can be understood that the second light source also has a structure similar to that of the first light source, which will not be repeated here.

In other embodiments, the first light source and the second light source can be disposed on the same base. As shown in FIG. 3 , the first light source 11 includes a plurality of lamp beads 111 . The second light source 12 includes a plurality of lamp beads 121 . Wherein, the first light source 11 and the second light source 12 can generate light of various colors according to different lamp beads.

The image acquisition component 20 is used to acquire the image to be inspected when the solder piece to be inspected is irradiated by the first color light and the second color light.

In some embodiments, the image acquisition component 20 may adopt a macro camera, so as to enlarge the solder ball on the solder piece to be inspected to a certain extent, so that the image to be inspected is clearer.

It can be understood that due to the improvement of the production process of the solder parts to be inspected, the more electronic components that can be welded on the solder parts to be inspected per unit area, the more corresponding solder balls, and the use of macro cameras for image collection can Acquired images are enlarged.

The processor 30 is connected with the image acquisition component 20, and is used for receiving the image to be inspected, and performing image recognition and analysis on the image to be inspected, so as to perform defect detection on the solder piece to be inspected. Wherein, in some embodiments, the processor 30 may be integrated with the image acquisition component 20 . In some other embodiments, the processor 30 may be connected with the image acquisition component 20 through a data line or a wireless communication connection.

In some embodiments, the processor 30 may use the defect detection model to perform image recognition and analysis on the image to be detected, so as to perform defect detection on the solder piece to be detected.

In an application scenario, the processor 30 may perform image recognition and analysis on the to-be-inspected image in the following manner, so as to perform defect detection on the to-be-inspected solder piece.

First, the processor 30 performs data preprocessing on the image to be inspected to detect whether the image size of the image to be inspected is a multiple of the step size in the defect detection model. Specifically, if the size of the image is not a multiple of the step size, the minimum multiple of the step size closest to the image size of the image to be detected is obtained.

In other embodiments, the image size may be padded so that the image size of the image to be detected is a multiple of the step size. In an application scenario, if the image to be detected is filled more, there will be information redundancy, which will affect the inference speed. Therefore, the method of reducing the black border can be used to improve the speed of reasoning. The least black border can be adaptively added to the image to be detected. For example, the 1000×800 image to be detected will not be directly scaled to the size of 608×608, but calculate 608/1000=0.608, then scale to the size of 608×486, then calculate 608-486=122, and then use the function np .mod(122, 32), take the remainder to get 26, where 32 is the step size. Then average 13 to fill the two ends of the height of the image to be detected, and finally the size of the image to be detected is 608×512.

Then, use the Focus module in the defect detection model to slice the image to be detected. The specific operation is to get a value for every other pixel in a picture, similar to neighboring downsampling, so that four pictures are obtained, the four pictures are complementary, and the length is about the same, but no information is lost. In this way, the The W and H information is concentrated in the channel space, and the input channel is expanded by 4 times, that is, the spliced picture becomes 12 channels compared with the original RGB three-channel mode, and finally the new picture obtained is subjected to a convolution operation, and finally A double downsampled feature map is obtained without information loss. For example, the image to be detected is input into the Focus module at 640×640×3, and the slice operation is used to first become a feature map of 320×320×12, and then undergo a convolution operation to finally become a feature map of 320×320×64. In this way, the feature extraction is more fully.

Then use the defect detection model to perform defect detection for each solder ball in the image to be detected, so as to obtain the defect type corresponding to each solder ball. For example, when the defect type is obtained, it can be marked in the original image and displayed, so that the user can clearly identify the defect solder ball, and facilitate subsequent maintenance. For example, re-soldering the solder piece to be tested, or scrapping the solder piece to be tested.

In some embodiments, a defect detection model may include an input layer, a feature extraction layer, a feature fusion layer, and an output layer.

Among them, data enhancement, adaptive anchor frame calculation, and adaptive image scaling can be performed on the image to be detected at the input layer. Among them, data enhancement can use mosaic data enhancement, by randomly scaling, randomly cropping, randomly arranging, and then splicing the images to be detected to obtain new images, which can improve the detection effect of small targets.

Among them, the adaptive anchor frame calculation, embeds this function into the code, and adaptively calculates the best anchor frame value in different training sets during each training.

Among them, adaptive image scaling reduces the black borders at both ends of the image height, and reduces the amount of calculation during inference, that is, the target detection speed will be improved.

In the feature extraction layer, the Focus layer, the first convolution layer, the first BottleneckCSP layer, the second convolution layer, the second BottleneckCSP layer, the third convolution layer, the third BottleneckCSP layer, the fourth convolution layer, SPP layer, the fourth BottleneckCSP layer.

Among them, the Focus layer is used to periodically extract pixels from the high-resolution image and reconstruct them into the low-resolution image, that is, to stack the four adjacent positions of the image, and focus the w and h dimension information into the channel space to improve Each point has a receptive field and reduces the loss of original information, which can reduce the amount of calculation and speed up the calculation.

Among them, the BottleneckCSP layer mainly includes a Bottleneck module and a CSP (Cross Stage Partial, cross-stage part) module.

Among them, the SPP layer adopts the maximum pooling of 5/9/13 respectively, and then performs concat fusion to improve the receptive field. The input of SPP is 512x20x20, after the 1x1 convolutional layer, the output is 256x20x20, and then three Maxpool (maximum pooling layers) are paralleled for downsampling, the result is added to the initial feature, and the output is 1024x20x20, and finally the convolution with 512 The core brings it back to 512x20x20.

The feature fusion layer adopts the structure of FPN (Feature Pyramid Networks, feature pyramid network) + PAN (Perceptual Adversarial Network, perceptual confrontation network), and adopts the CSP2 structure designed by referring to CSPnet (CrossStage Partial Networks, cross-stage local network) to strengthen network feature fusion Ability.

The output layer can use the GIOU_Loss function:

Specifically, the function formula is as follows:

Among them, IOU (Intersection over Union) means the ratio of intersection and union, and U means that it does not belong to the expression of the above formula. It means: first calculate the minimum closed area area Ac of the two frames (the area of the minimum frame that includes the predicted frame and the real frame at the same time ), and then calculate the IoU, and then calculate the proportion of the closure area that does not belong to the two boxes in the closure area Ac, and finally subtract this proportion from the IoU to obtain the GIoU.

Then the corresponding loss function is L _GIOU =1-GIOU.

It is also possible to use nms non-maximum suppression.

Use the above-mentioned solder ball defect detection device 100 to collect training images, and mark the solder balls in the training images, and then use the marked training images to train the above-mentioned defect detection model, so that after the training is completed, the processor 30. Using the defect detection model to perform image recognition and analysis on the image to be detected, so as to perform defect detection on the solder piece to be detected. It can be understood that using the training images collected by two different colors of light can make the solder ball defect in the image more obvious, easier to identify during labeling, and improve labeling efficiency. And because the solder ball defect in the image is more obvious, more features can be obtained during model training, so that the accuracy of the trained defect detection model can be improved, and then when using the defect detection model for defect detection, it can improve detection accuracy.

In other embodiments, the defect detection model may be trained using a target detection algorithm (yolov5 target detection algorithm), and the trained defect detection model may be used to perform solder ball defect detection on the image to be detected. Compared with traditional comparison and geometric calculation algorithms, this target detection algorithm can extract more features without manually designing a reasonable feature extraction method.

In the solder ball defect detection device provided in this embodiment, the solder ball defect detection device includes: a light source assembly, including a first light source and a second light source, the first light source is used to generate the first color light, and the second light source is used to generate The second color light, the first color light and the second color light are different in color, the first color light and the second color light are used to irradiate the solder piece to be detected; the image acquisition component is used to collect the solder piece to be detected by the first color The image to be inspected when the light and the second color light are irradiated; the processor is connected to the image acquisition component, and is used to receive the image to be inspected, and perform image recognition and analysis on the image to be inspected, so as to detect defects on the solder piece to be inspected. Through the above method, two different colors of light are used to irradiate the solder parts to be inspected, so that the images collected by the hemispherical solder balls on the solder parts to be inspected have obvious color differences, which can make the solder balls with defects more prominent. Further, the accuracy of the detection of the solder ball defect is improved, and the detection efficiency of the detection of the solder ball defect is improved.

Referring to FIG. 4 , FIG. 4 is a schematic structural diagram of an embodiment of a light source assembly and an image acquisition assembly provided by the present application. The light source assembly 10 is provided with an optical channel 14, and the image acquisition assembly 20 is arranged on the side of the light source assembly 10 away from the solder piece to be inspected. . The light source assembly 10 also includes a shield 13 for collecting the light generated by the light source assembly 10 so that the light generated by the light source assembly 10 can directly strike the solder piece to be inspected.

In some embodiments, referring to FIG. 5, the first light source 11 is provided with a first through hole 11a, the second light source 12 is provided with a second through hole 12a, and the first light source 11 and the second light source 12 are stacked so that The first through hole 11a communicates with the second through hole 12a to form the light channel 14; the diameter of the first through hole 11a is smaller than the diameter of the second through hole 12a. The image acquisition component 20 is disposed above the first light source 11 , that is, the side of the light source component 10 away from the solder piece to be inspected. The image acquisition component 20 performs image acquisition on the solder piece to be inspected through the first through hole 11a. In this way, the light generated by the first light source 11 is irradiated from the second through hole 12a to the solder piece to be inspected. The image acquisition component 20 performs image acquisition on the solder piece to be inspected through the optical channel 14 .

In some embodiments, referring to FIG. 6, the first light source 11 is provided with a first through hole 11a, the second light source is provided with a second through hole, and the second light source 12 is disposed in the first through hole 11a, so that the first The two through holes 12a form a light channel.

Referring to FIG. 7 , FIG. 7 is a schematic structural diagram of another embodiment of a solder ball defect detection device provided in the present application. The solder ball defect detection device 100 includes a light source assembly 10 , wherein the light source assembly 10 includes a first light source 11 , a second light source 12 , an image acquisition assembly 20 , a processor (not shown) and a support frame 40 .

Wherein, the first light source 11 , the second light source 12 and the image acquisition component 20 are connected to a power source 60 through a power line.

Further, the first light source 11 , the second light source 12 and the image acquisition component 20 are arranged on the support frame 40 . The supporting frame 40 is arranged on the storage table A, and the area B on the storage table A is used for placing the solder pieces to be inspected.

The first light source 11 , the second light source 12 , the image acquisition component 20 and the processor are the same as or similar to those in the above-mentioned embodiments, and will not be repeated here.

Referring to FIG. 8 , the support frame 40 includes a first support member 41 and a second support member 42 . The light source assembly 10 is arranged on the first support member 41, that is, the first light source 11 and the second light source 12 in the above embodiment are arranged on the first support member 41, and the image acquisition assembly 20 is arranged on the second support member 42; The second support 42 is disposed at one end of the first support 41 away from the optical paths of the first light source 11 and the second light source 12 , that is, the second support 42 is disposed at the side of the light source assembly 10 away from the solder piece to be inspected.

In some embodiments, the first light source 11 is disposed above the second light source 12 , and the light paths of the first light source 11 and the second light source 12 are directed toward the solder piece to be inspected.

Wherein, the first light source 11 includes a plurality of first sub-light sources, and the plurality of first sub-light sources are arranged in a ring to form a first through hole 11a in the plurality of first sub-light sources; the second light source 12 includes a plurality of second sub-light sources A plurality of second sub-light sources are arranged in a ring to form second through holes 12a in the plurality of second sub-light sources.

The first light source 11 and the second light source 12 are annular, and the image acquisition component 20 is arranged above the first light source 11, and the image acquisition component 20 passes through the first through hole 11a and the second through hole 11a of the first light source 11 and the second light source 12. The through hole 12a performs image acquisition on the solder piece to be inspected to obtain an image to be inspected. The first light source 11 on the top will form a small dot on the qualified tin ball top in the solder piece to be detected, and the second light source 12 below will illuminate around the solder ball of the solder piece to be detected, and the defect in the solder piece to be detected The top of the ball will be irregular in shape, oversaturated (dark color), etc.

In some embodiments, the first light source 11 and the second light source 12 can be arranged on the same substrate, for example, the first light source 11 and the second light source 12 are respectively composed of a plurality of LED lamp beads, wherein the plurality of LED lamp beads according to A certain number of groups, therefore, the first light source 11 and the second light source 12 are respectively composed of multiple groups of lamp beads. Referring to FIG. 9 , the first light source 11 is composed of a plurality of lamp beads D, and the second light source 12 is composed of a plurality of lamp beads E. The lamp bead D and the lamp bead E are arranged on the substrate at intervals. Wherein, a through hole C is formed in the central area of the substrate, and the image acquisition component 20 can collect the image to be detected through the through hole C. The irradiation ranges of the first light source 11 and the second light source 12 can cover the solder parts to be inspected.

In any of the above embodiments, the first color light emitted by the first light source 11 may be monochromatic light, and the second color light emitted by the second light source 12 may be polychromatic light or monochromatic light. Alternatively, the first color light emitted by the first light source 11 may be polychromatic light or monochromatic light, and the second color light emitted by the second light source 12 may be monochromatic light.

For example, the first color light may be red light, and the second color light may be white light. In another example, the first color light may be red light, and the second color light may be blue light.

In an application scenario, refer to FIG. 10 , which is a schematic diagram of an image to be detected provided by the present application after defect detection. It can be understood that the solder ball is semicircular after correct soldering, therefore, the image of a defect-free solder ball will have a dot at the apex of the solder ball. The usual defect types of defective solder balls are stain defects, crush defects, crack defects, and irregular shiny defects. The solder ball with defects is not a complete semicircle, therefore, the first color light will form a defect area corresponding to the defect on the surface of the solder ball. As shown in FIG. 10 , defective solder balls exist in regions G and H of the image F to be detected. It can be understood that in other embodiments, the defect types of defective solder balls can be displayed correspondingly. It can be understood that when labeling defects on training images, the above-mentioned defect types are also used for labeling.

Through the above method, two different colors of light are used to irradiate the solder parts to be inspected, so that the images collected by the hemispherical solder balls on the solder parts to be inspected have obvious color differences, which can make the solder balls with defects more prominent. Further, the accuracy of the detection of the solder ball defect is improved, and the detection efficiency of the detection of the solder ball defect is improved.

Referring to FIG. 11 , FIG. 11 is a schematic flowchart of an embodiment of a solder ball defect detection method provided by the present application. The method includes:

Step 81: Obtain an image to be detected.

Wherein, the image to be inspected is acquired by using the image acquisition component 20 in the solder ball defect inspection device 100 in any of the above-mentioned embodiments when the solder piece to be inspected is irradiated by light of two different colors. Specifically, the solder piece to be inspected is irradiated by the light of the first color emitted by the first light source and the light of the second color emitted by the second light source.

Step 82: Use the defect detection model to detect the image to be detected, and obtain the detection information of the corresponding defect in the image to be detected.

Wherein, the defect detection model is obtained by training with training images, wherein, the training images are collected by the image acquisition component 20 in the solder ball defect detection device 100 in any of the above-mentioned embodiments when the solder piece to be detected is irradiated by two different colors of light get. Specifically, the solder piece to be inspected is irradiated by the light of the first color emitted by the first light source and the light of the second color emitted by the second light source. After obtaining the training image, mark it according to the actual defect of the tin ball on the training image, and then train the defect detection model according to the real information of the mark.

Through the above method, two different colors of light are used to irradiate the solder parts to be inspected, so that the images collected by the hemispherical solder balls on the solder parts to be inspected have obvious color differences, which can make the solder balls with defects more prominent. Further, the accuracy of the detection of the solder ball defect is improved, and the detection efficiency of the detection of the solder ball defect is improved.

Referring to FIG. 12, FIG. 12 is a schematic structural diagram of an embodiment of an electronic device provided by the present application. The electronic device 90 includes a controller 91 and a memory 92 connected to the controller 91; wherein, the memory 92 is used to store program data, and the controller 91 Used to execute program data to implement the following methods:

Obtain the image to be detected; use the defect detection model to detect the image to be detected, and obtain the detection information of the corresponding defect in the image to be detected, wherein the defect detection model is obtained by training the training image, wherein the image to be detected and the training image are obtained using the above-mentioned technology The image acquisition component in the solder ball defect detection device provided by the solution is acquired when the solder piece to be inspected is irradiated by two different colors of light.

It can be understood that the controller 91 may be the processor 30 in the solder ball defect detection device 100 of any of the above-mentioned embodiments. For specific implementation manners, reference may be made to the foregoing embodiments, and details are not repeated here.

Referring to FIG. 13, FIG. 13 is a schematic structural diagram of an embodiment of a computer-readable storage medium provided by the present application. The computer-readable storage medium 130 is used to store program data 131. When the program data 131 is executed by a processor, it is used to realize The following method steps:

Obtain the image to be detected; use the defect detection model to detect the image to be detected, and obtain the detection information of the corresponding defect in the image to be detected, wherein the defect detection model is obtained by training the training image, wherein the image to be detected and the training image are obtained using the above-mentioned technology The image acquisition component in the solder ball defect detection device provided by the solution is acquired when the solder piece to be inspected is irradiated by two different colors of light.

It can be understood that the computer-readable storage medium 130 in this embodiment is applied to an electronic device, and its specific implementation steps can refer to the above-mentioned embodiments, which will not be repeated here.

In the above-mentioned solder ball defect detection device, solder ball defect detection method, electronic equipment, and computer-readable storage medium of the present application, the images to be detected are collected in the form of irradiating the solder parts to be detected with two different colors of light, and the images to be detected are collected in two colors The light with obviously different hues illuminates the solder parts to be inspected, which can effectively distinguish the defective solder balls and qualified solder balls in the solder parts to be inspected. Secondly, using the target detection algorithm, the solder parts to be detected only need to appear in the field of view of the image acquisition component, and learn effective feature extraction from big data. The target detection algorithm can be accelerated by special acceleration hardware to achieve more efficient detection of solder balls defect.

In the several implementation manners provided in this application, it should be understood that the disclosed methods and devices may be implemented in other ways. For example, the device implementation described above is only illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be Incorporation may either be integrated into another system, or some features may be omitted, or not implemented.

The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

If the integrated units in the above other embodiments are realized in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or part of the contribution to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , including several instructions to make a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (processor) execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk, and other media that can store program codes.

The above is only the implementation of the application, and does not limit the patent scope of the application. Any equivalent structure or equivalent process conversion made by using the specification and drawings of the application, or directly or indirectly used in other related technologies fields, are all included in the scope of patent protection of this application in the same way.

Claims (12)

1. A solder ball defect detecting apparatus, comprising:

the light source assembly comprises a first light source and a second light source, wherein the first light source is used for generating first color light, the second light source is used for generating second color light, the colors of the first color light and the second color light are different, and the first color light and the second color light are used for irradiating the soldering tin piece to be detected;

the image acquisition assembly is used for acquiring an image to be detected when the soldering tin piece to be detected is irradiated by the first color light and the second color light;

and the processor is connected with the image acquisition assembly and is used for receiving the image to be detected and carrying out image recognition analysis on the image to be detected so as to detect the defects of the soldering tin piece to be detected.

2. The solder ball defect detecting apparatus of claim 1,

the light source assembly is provided with a light channel, the image acquisition assembly is arranged on one side, away from the soldering tin piece to be detected, of the light source assembly, and light generated by the light source assembly enters the image acquisition assembly through the light channel after being reflected by the soldering tin piece to be detected.

3. The solder ball defect detecting apparatus of claim 2,

the first light source is provided with a first through hole, the second light source is provided with a second through hole, and the first light source and the second light source are arranged in a stacked mode so that the first through hole and the second through hole are communicated to form the light channel;

the aperture of the first through hole is smaller than the aperture of the second through hole.

4. The solder ball defect detecting apparatus of claim 2,

the first light source is provided with a first through hole, the second light source is provided with a second through hole, and the second light source is arranged in the first through hole so that the second through hole forms the optical channel.

5. The solder ball defect detecting apparatus of claim 3 or 4,

the first light source comprises a plurality of first sub-light sources which are arranged in a ring shape to form the first through holes in the plurality of first sub-light sources;

the second light source includes a plurality of second sub light sources arranged in a ring shape to form the second through holes in the plurality of second sub light sources.

6. The solder ball defect detecting apparatus of claim 1,

the tin ball defect detection device further comprises a support frame, and the light source assembly and the image acquisition assembly are arranged on the support frame.

7. The solder ball defect detecting apparatus of claim 6,

the support frame comprises a first support part and a second support part; the light source assembly is arranged on the first supporting piece, and the image acquisition assembly is arranged on the second supporting piece; the second supporting piece is arranged on one side, away from the soldering tin piece to be detected, of the light source component.

8. The solder ball defect detecting apparatus of claim 1,

the first color light is monochromatic light, and the second color light is polychromatic light or monochromatic light.

9. The solder ball defect detecting apparatus of claim 8,

the first color light is red light, and the second color light is white light.

10. A solder ball defect detection method is characterized by comprising the following steps:

acquiring an image to be detected;

detecting the image to be detected by using a defect detection model to obtain detection information of the corresponding defect in the image to be detected, wherein the defect detection model is obtained by training a training image, and the image to be detected and the training image are obtained by using an image acquisition component in the solder ball defect detection device as claimed in any one of claims 1 to 9 when the solder ball to be detected is irradiated by light with two different colors.

11. An electronic device, comprising a controller and a memory to which the controller is connected;

wherein the memory is configured to store program data and the controller is configured to execute the program data to implement the method of claim 10.

12. A computer-readable storage medium, characterized in that the computer-readable storage medium is used for storing program data, which, when being executed by a processor, is used for carrying out the method as claimed in claim 10.

Images