CN105354816B - Electronic component positioning method and device - Google Patents

Abstract

The invention discloses a method for locating electronic components, comprising the following steps: performing background modeling on at least two images of the front panel of the plug-in that are collected, and obtaining a model of each pixel of the background model; K probability values of each pixel point under the k Gaussian distribution functions of the corresponding pixel points on the background model; the k probability values are compared with a preset threshold one by one, and at any probability When the value is less than the threshold, the corresponding pixel is marked as a candidate component pixel on the plug-in rear panel image; the adjacent candidate component pixels are connected on the plug-in rear panel image to form at least one connected region to locate the electronic components. The invention also discloses an electronic component positioning device. The invention realizes fast positioning of electronic components, and improves the production speed of the standard layout in the circuit board detection process.

Description

Method and device for positioning electronic components

technical field

The invention relates to the field of automatic detection, in particular to an electronic component positioning method and device.

Background technique

Automatic optical inspection refers to the use of optical imaging to obtain the surface state of the finished product, and to detect whether there are foreign objects or surface defects on the surface of the finished product through image processing. At present, automatic optical inspection is widely used in the quality inspection of circuit boards. During detection, the relevant detection device automatically scans the circuit board through the camera to obtain images, extracts partial images of each electronic component, and uses image processing technology to determine whether the electronic components on the circuit board have defects such as wrong insertion, missing insertion or reverse insertion. , and finally display or mark the suspected defective electronic components for easy viewing and maintenance.

Before detecting defects in electronic components, it is necessary to make a standard layout of the circuit board. In particular, it is necessary to mark the position of each electronic component on the circuit board. The existing solution is to set the position of each electronic component on the circuit board by manual operation. However, when the number of electronic components is large, the manual operation is not only time-consuming, but also prone to missing electronic components. To meet the needs of use.

Contents of the invention

In view of the above problems, the object of the present invention is to provide a method and device for locating electronic components, which can quickly and accurately locate the positions of all electronic components on the image of the circuit board.

An embodiment of the present invention provides a method for locating electronic components, including the following steps:

Background modeling is performed on at least two images of the front panel of the plug-in to obtain a model of each pixel of the background model, wherein the image of the front panel of the plug-in is an image of a circuit board without electronic components inserted, and the model of each pixel is The model consists of k Gaussian distribution functions, k is an integer greater than 1;

Calculating the k probability values of each pixel of the collected plug-in back panel image under the k Gaussian distribution functions of the corresponding pixels on the background model, wherein the plug-in back panel image is an image of the inserted electronic component image of the circuit board;

Comparing the k probability values with a preset threshold value one by one, and when any probability value is smaller than the threshold value, marking the corresponding pixel on the plug-in rear panel image as a candidate component pixel;

Connect adjacent candidate component pixels on the image of the plug-in backboard to form at least one connected region to locate the electronic component.

As an improvement of the above scheme, the Gaussian mixture model is used to carry out background modeling on at least two plug-in front panel images collected, and obtain a model of each pixel of the background model obtained according to the background modeling, including:

Build a model p(x) for each pixel in any of the front panel images of the plug-in, where, x is the gray value of the pixel, k is the number of Gaussian models, ω _j , μ _j , and C _j represent the weight, mean and covariance of the jth Gaussian model respectively;

The weight, mean value and covariance of the established model of each pixel point are updated by using the corresponding pixel points on the front panel image of other plug-ins to obtain the updated model p(x) of each pixel point.

As an improvement of the above scheme, the k probability values under the k Gaussian distribution functions of the corresponding pixels of each pixel of the collected plug-in rear plate image on the background model are calculated respectively, specifically:

Calculate the probability value p _j (y) of each pixel point y on the image of the plug-in rear panel under the k Gaussian distribution functions of the corresponding pixel points of the background model, wherein, And 1≤j≤k.

As an improvement of the above solution, the connected area is rectangular.

As an improvement of the above solution, adjacent candidate component pixels are connected on the image of the plug-in rear board to form at least one connected region to locate the region where the electronic component is located, including:

calculating the area of the at least one connected region;

Determine whether the area of each connected region is greater than a preset area threshold;

When the area of the connected area is greater than the area threshold, mark the connected area as an effective area containing electronic components to locate the electronic components; otherwise, mark the connected area as an interference area.

An embodiment of the present invention also provides an electronic component positioning device, including:

A modeling unit, configured to perform background modeling on at least two images of the front board of the plug-in, and obtain a model of each pixel of the background model, wherein the image of the front board of the plug-in is an image of a circuit board without electronic components inserted , the model of each pixel is composed of k Gaussian distribution functions, k is an integer greater than 1;

A probability value calculation unit, configured to separately calculate k probability values under k Gaussian distribution functions of k Gaussian distribution functions corresponding to pixels on the background model for each pixel of the collected plug-in rear panel image, wherein, the plug-in rear The board image is an image of a circuit board into which electronic components are inserted;

A comparing unit, configured to compare the k probability values with a preset threshold one by one, and when any probability value is smaller than the threshold, mark the corresponding pixel on the plug-in rear panel image as a candidate component pixel;

The positioning unit is configured to connect adjacent candidate component pixels on the image of the plug-in backboard to form at least one connected area, so as to locate the area where the electronic component is located.

As an improvement of the above scheme, the modeling unit includes:

The model building unit is used to set up a model p(x) for each pixel in any one plug-in front panel image, wherein, x is the gray value of the pixel, k is the number of Gaussian models, ω _j , μ _j , and C _j represent the weight, mean and covariance of the jth Gaussian model respectively;

The update unit is configured to update the weight, mean value and covariance of the established model of each pixel by using the corresponding pixels on the front panel image of other plug-ins to obtain the updated model of each pixel.

As an improvement to the above scheme, the probability value calculation unit is specifically configured to calculate the probability of each pixel point y on the plug-in rear panel image under the k Gaussian distribution functions of the corresponding pixel points of the background model value p _j (y), where, And 1≤j≤k.

As an improvement of the above solution, the connected area is rectangular.

As an improvement of the above solution, the positioning unit includes:

an area calculation unit, configured to calculate the area of the at least one connected region;

A judging unit, used for whether the area of each connected region is greater than a preset area threshold;

A marking unit, configured to mark the connected area as an effective area containing electronic components when the area of the connected area is greater than the area threshold, so as to locate the electronic components; otherwise, mark the connected area as an interference area .

The electronic component positioning method and device provided by the embodiments of the present invention establish a background model by using a Gaussian mixture model, and then match each pixel point of the background model according to the image of the rear panel of the plug-in, and obtain a candidate according to the matching situation. Component pixels, and by connecting the adjacent candidate component pixels, locate the position of the electronic component on the image of the plug-in back board, thereby realizing the fast and accurate positioning of the position of the electronic component from the image of the plug-in back board, for the subsequent Board inspection provides a reliable standard format.

Description of drawings

In order to illustrate the technical solution of the present invention more clearly, the accompanying drawings used in the implementation will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some implementations of the present invention. As far as the skilled person is concerned, other drawings can also be obtained based on these drawings on the premise of not paying creative work.

FIG. 1 is a flowchart of an electronic component positioning method provided by an embodiment of the present invention.

Fig. 2 is a schematic diagram of a plug-in front panel provided by an embodiment of the present invention.

Fig. 3 is a schematic diagram of a plug-in backplane provided by an embodiment of the present invention.

Fig. 4 is a schematic diagram of locating electronic components in the image of the plug-in backboard provided by an embodiment of the present invention.

Fig. 5 is a flowchart of an electronic component positioning device provided by an embodiment of the present invention.

FIG. 6 is a schematic structural diagram of the modeling unit shown in FIG. 5 .

FIG. 7 is another structural schematic diagram of the positioning unit shown in FIG. 5 .

detailed description

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

Embodiments of the present invention provide a method and device for locating electronic components, which are used for locating the positions of all electronic components on a circuit board through automatic positioning. Detailed descriptions are given below respectively.

Please refer to FIG. 1 . FIG. 1 is a flowchart of an electronic component positioning method provided by an embodiment of the present invention. The electronic component locating method can be executed by an electronic component locating device, and at least includes steps S101 to S104. in,

S101. Perform background modeling on at least two images of the front board of the plug-in, and obtain a model of each pixel of the background model, wherein the front board image of the plug-in is an image of a circuit board without electronic components inserted, and each pixel The point model consists of k Gaussian distribution functions, where k is an integer greater than 1.

Please refer to FIG. 2 together. In the embodiment of the present invention, the image of the front board of the plug-in is an image of a circuit board without electronic components inserted therein. Wherein, the electronic component positioning device may use a Gaussian mixture model to model the image of the plug-in front board to obtain a background model of the plug-in front board. When the Gaussian mixture model performs background modeling, in order to describe the background and its possible changes, multiple background images are required. Therefore, there are at least two images of the plug-in front panel in the present invention.

Specifically, when performing background modeling:

First, establish a model p(x) for each pixel in any image of the front panel of the plug-in, where, x is the gray value of the pixel, k is the number of Gaussian models, ω _j , μ _j , C _j respectively represent the weight, mean and covariance of the jth Gaussian model.

In the embodiment of the present invention, the electronic component positioning device randomly takes an image of the front panel of the plug-in, and establishes a model p(x) for each pixel of the image of the front panel of the plug-in, wherein, x is the pixel value of the pixel point, k is the number of Gaussian models, ω _j , μ _j , C _j respectively represent the weight, mean and covariance of the jth Gaussian model. Here, it can be considered that the pixel value of one pixel of the front panel image of the plug-in is expressed as a combination of k Gaussian distribution functions, and the weight, mean value and covariance of the Gaussian model can be set according to empirical values first.

Then, the weight, mean and covariance of the established model of each pixel point are updated by using the corresponding pixels on the front panel image of other plug-ins to obtain the updated model p(x) of each pixel point.

In the embodiment of the present invention, when a new pixel is added (that is, a new plug-in front panel image is added), the electronic component positioning device compares the pixel value of the new pixel with the pixel value of the corresponding background model Compared with the mean μ _j of the k Gaussian distributions of the point, the probability of the new pixel value falling into the corresponding Gaussian distribution is calculated at the same time, and the matching Gaussian distribution is selected according to the judgment rule. When there is a matching Gaussian distribution, it is necessary to update the parameters of these Gaussian distributions such as weight, mean and covariance according to the pixel value of the new pixel. For the specific calculation process, reference may be made to the principle of updating parameters of a general Gaussian mixture model, which will not be described in detail here.

In the embodiment of the present invention, the electronic component positioning device obtains the model of each pixel of the background model through the above operations, wherein the model of each pixel is composed of k Gaussian distribution functions, and k is greater than 1 is an integer, and preferably, the value range of k is 3 to 5.

S102, respectively calculate k probability values of each pixel of the collected plug-in rear panel image under the k Gaussian distribution functions of the corresponding pixels on the background model, wherein the plug-in rear panel image is the plug-in electronic An image of a circuit board with components.

Please also refer to FIG. 3 , in the embodiment of the present invention, the electronic component positioning device first obtains an image of the back board of the plug-in, wherein the image of the back board of the plug-in is an image of a circuit board where electronic components are inserted. Then, the electronic component positioning device respectively calculates k probability values of each pixel of the plug-in rear board image under k Gaussian distribution functions of corresponding pixels on the background model.

For example, it is assumed that the size of the background template and the plug-in back panel image are both M×N pixels. Wherein, each pixel of the background template can be represented by k Gaussian distribution functions. When performing calculations, the electronic component positioning device first reads the pixel value y of the first pixel point (such as coordinates (1,1)) of the image of the back plate of the plug-in, and then the electronic component positioning device reads k Gaussian distribution functions of pixels on the corresponding background model (for example, the coordinates are also (1,1)) Then, the electronic component positioning device respectively substitutes the pixel value y of the image of the plug-in back plate into the k Gaussian distribution functions to obtain k probability values, wherein the jth probability value can be expressed as

In this way, after the electronic component locating device traverses all the pixels on the image of the plug-in rear board and the background template, k probability values corresponding to each pixel can be obtained.

S103, comparing the k probability values with a preset threshold one by one, and when any probability value is smaller than the threshold, mark the corresponding pixel on the plug-in rear panel image as a candidate component pixel.

In the embodiment of the present invention, the electronic component locating device compares the k probability values with a preset threshold value one by one, wherein, when any probability value is smaller than the threshold value, the corresponding pixel point mark as a candidate component pixel, and when all the k probability values are greater than the threshold, then mark the pixel as a background pixel.

S104. Connect adjacent candidate component pixels on the plug-in backboard image to form at least one connected region, so as to locate the electronic component.

Please refer to FIG. 4 together. In one embodiment of the present invention, the electronic component locating device connects adjacent candidate component pixels after marking all the pixels on the image of the plug-in rear board. Wherein, after being connected, at least one connected area will be formed, and the connected area is where the electronic components are located. Wherein, preferably, as a preferred embodiment of the present invention, the shape of the connected region is a rectangle (as shown by three white boxes in FIG. 4 ). Of course, it can be understood that, in other embodiments of the present invention, the communication area may also be circular, triangular or other shapes, which are not specifically limited in the present invention.

In another embodiment of the present invention, due to problems such as the accuracy of the algorithm or parameter errors, sometimes pixels that were originally background pixels are miscalculated as candidate component pixels. At this time, a connection area may be formed at a position where no electronic component originally exists. In order to prevent the above situation, the electronic component positioning device can be set as follows:

First, the area of the at least one connected region is calculated.

Secondly, it is judged whether the area of each connected region is greater than a preset area threshold.

Finally, when the area of the connected region is greater than the area threshold, mark the connected region as an effective region containing electronic components; otherwise, mark the connected region as an interference region not containing electronic components.

In this way, those connected areas with smaller areas are excluded, that is, those interference areas caused by calculation accuracy or errors are excluded, so as to ensure that the connected areas are all areas including electronic components.

To sum up, the electronic component positioning method provided by the embodiment of the present invention uses a Gaussian mixture model to establish a background model, and then matches each pixel of the background model according to the picture of the plug-in rear board to obtain candidate component pixels , and by connecting the pixels of adjacent candidate components, the position of the electronic component is located on the image of the back board of the plug-in, so as to realize the fast and accurate positioning of the position of the electronic component from the image of the back board of the plug-in, and provide for the subsequent circuit board detection Solid standard layout.

Please refer to FIG. 5 . FIG. 5 is a structural diagram of an electronic component positioning device provided by an embodiment of the present invention. The electronic component locating device 100 can be used to implement the above-mentioned electronic component locating method, and at least includes a modeling unit 10, a probability value calculation unit 20, a comparing unit 30 and a locating unit 40, wherein:

The modeling unit 10 is configured to perform background modeling on the collected at least two plug-in front board images to obtain a model of each pixel of the background model, wherein the plug-in front board images are circuits without electronic components inserted The image of the plate, the model of each pixel is composed of k Gaussian distribution functions, k is an integer greater than 1.

In the embodiment of the present invention, the image of the front board of the plug-in is an image of a circuit board without electronic components inserted therein. In order to locate the position of the electronic components from the circuit board, the modeling unit 10 may first collect at least two images of the front board of the plug-in as background images. Wherein, the modeling unit 10 may use a Gaussian mixture model to model the image of the plug-in front board to obtain a background model of the plug-in front board. When the Gaussian mixture model performs background modeling, in order to describe the background and its possible changes, multiple background images are required. Therefore, there are at least two images of the plug-in front panel in the present invention.

Please refer to FIG. 6 together. Specifically, the modeling unit 10 includes a model building unit 11 and an updating unit 12, wherein:

The model building unit 11 is configured to set up a model p(x) for each pixel in any image of the front panel of the plug-in, wherein, x is the gray value of the pixel, a is the number of Gaussian models, ω _j , μ _j , C _j respectively represent the weight, mean and covariance of the jth Gaussian model.

In the embodiment of the present invention, the model building unit 11 randomly takes an image of the front panel of the plug-in, and establishes a model p(x) for each pixel of the image of the front panel of the plug-in, wherein, x is the pixel value of the pixel point, k is the number of Gaussian models, ω _j , μ _j , C _j respectively represent the weight, mean and covariance of the jth Gaussian model. Here, it can be considered that the pixel value of one pixel of the front panel image of the plug-in is expressed as a combination of k Gaussian distribution functions, and the weight, mean value and covariance of the Gaussian model can be set according to empirical values first.

The update unit 12 is used to update the weight, mean and covariance of the established model of each pixel point by using the corresponding pixels on the front panel image of other plug-ins to obtain the updated model of each pixel point p(x).

In the embodiment of the present invention, when a new pixel is added (that is, a new plug-in front panel image is added), the update unit 12 compares the pixel value of the new pixel with the above-mentioned k Gaussian distribution At the same time, the probability that the new pixel value falls into the corresponding _Gaussian distribution is calculated, and the matching Gaussian distribution is selected according to the judgment rule. When there is a matching Gaussian distribution, it is necessary to update the parameters of these Gaussian distributions such as weight, mean and covariance according to the pixel value of the new pixel. For the specific calculation process, reference may be made to the principle of updating parameters of a general Gaussian mixture model, which will not be described in detail here.

In the embodiment of the present invention, the modeling unit 10 obtains the model of each pixel of the background model through the above operations, wherein the model of each pixel is composed of k Gaussian distribution functions, and k is greater than 1 is an integer, and preferably, the value range of k is 3 to 5.

The probability value calculation unit 20 is used to separately calculate k probability values under k Gaussian distribution functions of k Gaussian distribution functions of each pixel point of the collected plug-in rear plate image on the corresponding pixel point on the background model, wherein, The aforementioned image of the backboard of the plug-in is an image of the circuit board on which the electronic components are inserted.

In the embodiment of the present invention, the probability value calculation unit 20 first acquires an image of a plug-in back board, wherein the image of the plug-in back board is an image of a circuit board where electronic components are inserted. Then, the probability value calculation unit 20 respectively calculates k probability values of each pixel of the plug-in rear panel image under the k Gaussian distribution functions of the corresponding pixel on the background model.

For example, it is assumed that the size of the background template and the plug-in back panel image are both M×N pixels. Wherein, each pixel of the background template can be represented by k Gaussian distribution functions. When calculating, the probability value calculation unit 20 first reads the pixel value y of the first pixel point (such as coordinates (1,1)) of the plug-in back plate image, and then the probability value calculation unit 20 Read the k Gaussian distribution functions of the pixels on the corresponding background model (for example, the coordinates are also (1,1)) Then, the probability value calculation unit 20 substitutes the pixel value y distribution of the plug-in rear panel image into the k Gaussian distribution functions to obtain k probability values, wherein the jth probability value can be expressed as

In this way, after the probability value calculation unit 20 traverses all the pixels on the plug-in back panel image and the background template, k probability values corresponding to each pixel can be obtained.

The comparison unit 30 is configured to compare the k probability values with a preset threshold value one by one, and when any probability value is smaller than the threshold value, the corresponding pixel point will be displayed on the image of the plug-in rear panel Mark the pixels as candidate components.

In the embodiment of the present invention, the comparison unit 30 compares the k probability values with a preset threshold one by one, wherein, when any probability value is smaller than the threshold, the comparison unit 30 will The corresponding pixel is marked as a candidate component pixel, and when all the k probability values are greater than the threshold, the comparison unit 30 marks the pixel as a background pixel.

The positioning unit 40 is configured to connect adjacent candidate component pixels on the image of the plug-in backboard to form at least one connected area for locating the electronic component.

In an embodiment of the present invention, the positioning unit 40 connects adjacent candidate component pixels after marking all the pixel points on the image of the plug-in rear board. Wherein, after being connected, at least one connected area will be formed, and the connected area is where the electronic components are located. Wherein, preferably, as a preferred embodiment of the present invention, the shape of the communication area is a rectangle. Of course, it can be understood that, in other embodiments of the present invention, the communication area may also be circular, triangular or other shapes, which are not specifically limited in the present invention.

Please also refer to FIG. 7 . In another embodiment of the present invention, due to problems such as algorithm accuracy or parameter errors, sometimes pixels that were originally background pixels are miscalculated as candidate component pixels. At this time, a connection area may be formed at a position where no electronic component originally exists. In order to prevent the above situation, the positioning unit 40 may include:

An area calculation unit 41, configured to calculate the area of the at least one connected region.

A judging unit 42, configured to judge whether the area of each connected region is greater than a preset area threshold.

A marking unit 43, configured to mark the connected region as an effective region containing electronic components when the area of the connected region is greater than the area threshold; otherwise, mark the connected region as an interference region not containing electronic components .

In this way, the positioning unit 40 excludes those connected areas with small areas, that is, excludes those interference areas caused by calculation accuracy or errors, so as to ensure that the connected areas are all areas including electronic components.

To sum up, the electronic component locating device 100 provided by the embodiment of the present invention establishes a background model by using a Gaussian mixture model, and then matches each pixel of the background model according to the picture of the plug-in rear board to obtain a candidate component pixels, and by connecting adjacent candidate component pixels, locate the position of the electronic component on the image of the back board of the plug-in, so as to realize the fast and accurate positioning of the position of the electronic component from the image of the back board of the plug-in, for the subsequent circuit board inspection Provide reliable standard layout.

What is disclosed above is only a preferred embodiment of the present invention, and of course it cannot limit the scope of rights of the present invention. Those of ordinary skill in the art can understand all or part of the process for realizing the above embodiments, and according to the rights of the present invention The equivalent changes required still belong to the scope covered by the invention.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented through computer programs to instruct related hardware, and the programs can be stored in a computer-readable storage medium. During execution, it may include the processes of the embodiments of the above-mentioned methods. Wherein, the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random Access Memory, RAM) and the like.

Claims (10)

1. An electronic component positioning method, is characterized in that, comprises the steps: Background modeling is performed on at least two images of the front panel of the plug-in to obtain a model of each pixel of the background model, wherein the image of the front panel of the plug-in is an image of a circuit board without electronic components inserted, and the model of each pixel is The model consists of k Gaussian distribution functions, k is an integer greater than 1; Calculating the k probability values of each pixel of the collected plug-in back panel image under the k Gaussian distribution functions of the corresponding pixels on the background model, wherein the plug-in back panel image is an image of the inserted electronic component image of the circuit board; Comparing the k probability values with a preset threshold value one by one, and when any probability value is smaller than the threshold value, marking the corresponding pixel on the plug-in rear panel image as a candidate component pixel; Connect adjacent candidate component pixels on the image of the plug-in backboard to form at least one connected region to locate the electronic component. 2. The electronic component positioning method according to claim 1, wherein the background modeling is carried out on the at least two plug-in front panel images collected, and the model of each pixel of the background model is obtained, including: Build a model p(x) for each pixel in any of the front panel images of the plug-in, where, x is the gray value of the pixel, k is the number of Gaussian models, ω _j , μ _j , and C _j represent the weight, mean and covariance of the jth Gaussian model respectively; The weight, mean value and covariance of the established model of each pixel point are updated by using the corresponding pixel points on the front panel image of other plug-ins to obtain the updated model p(x) of each pixel point. 3. The electronic component positioning method according to claim 2, characterized in that, k Gaussian distribution functions of each pixel of the plug-in rear plate image collected by the calculation respectively on the corresponding pixels on the background model The following k probability values, specifically: Calculate the probability value p _j (y) of each pixel point y on the image of the plug-in rear panel under the k Gaussian distribution functions of the corresponding pixel points of the background model, wherein, And 1≤j≤k. 4. The method for locating electronic components according to claim 1, wherein the connected area is rectangular. 5. The electronic component positioning method according to any one of claims 1 to 4, wherein adjacent candidate component pixels are connected on the image of the plug-in rear board to form at least one connected area to locate the electronic component components, including: calculating the area of the at least one connected region; Determine whether the area of each connected region is greater than a preset area threshold; When the area of the connected area is greater than the area threshold, mark the connected area as an effective area containing electronic components to locate the electronic components; otherwise, mark the connected area as an interference area. 6. An electronic component positioning device, characterized in that it comprises: A modeling unit, configured to perform background modeling on at least two images of the front board of the plug-in, and obtain a model of each pixel of the background model, wherein the image of the front board of the plug-in is an image of a circuit board without electronic components inserted , the model of each pixel is composed of k Gaussian distribution functions, k is an integer greater than 1; A probability value calculation unit, configured to separately calculate k probability values under k Gaussian distribution functions of k Gaussian distribution functions corresponding to pixels on the background model for each pixel of the collected plug-in rear panel image, wherein, the plug-in rear The board image is an image of a circuit board into which electronic components are inserted; A comparing unit, configured to compare the k probability values with a preset threshold one by one, and when any probability value is smaller than the threshold, mark the corresponding pixel on the plug-in rear panel image as a candidate component pixel; A positioning unit, configured to connect adjacent candidate component pixels on the image of the plug-in backboard to form at least one connected region to locate the electronic component. 7. The electronic component positioning device according to claim 6, wherein the modeling unit comprises: The model building unit is used to set up a model p(x) for each pixel in any one plug-in front panel image, wherein, x is the gray value of the pixel, k is the number of Gaussian models, ω _j , μ _j , and C _j represent the weight, mean and covariance of the jth Gaussian model respectively; The update unit is used to update the weight, mean and covariance of the established model of each pixel by using the corresponding pixels on the front panel image of other plug-ins to obtain the updated model p(x ). 8. The electronic component locating device according to claim 7, wherein the probability value calculation unit is specifically used to calculate the correspondence of each pixel point y on the image of the rear board of the plug-in to the background model The probability value p _j (y) under the k Gaussian distribution functions of the pixel point, where, And 1≤j≤k. 9. The electronic component locating device according to claim 6, wherein the communication area is rectangular. 10. The electronic component positioning device according to any one of claims 6 to 9, wherein the positioning unit comprises: an area calculation unit, configured to calculate the area of the at least one connected region; A judging unit, used for whether the area of each connected region is greater than a preset area threshold; A marking unit, configured to mark the connected region as an effective region containing electronic components when the area of the connected region is greater than the area threshold, so as to locate the electronic component; otherwise, mark the connected region as an interference region .