CN118097092A - A method, device and system for intelligent inspection of the crimping quality of micro pins of electrical connectors - Google Patents

Abstract

The invention relates to the technical field of information, and discloses an intelligent checking method, device and system for the press-connection quality of a miniature contact pin of an electric connector, wherein the device comprises the following components: the box body, the box cover and the camera device; the method comprises the following steps: s1, off-line modeling, namely detecting and positioning key parts of the contact pin in real time by using a high-quality and precisely marked image data set through a deep learning training model; s2, checking the press-connection quality of the pin, and checking the state of the in-pin pressure and the conformity of the exposed length of the wire core; s3, displaying an inspection result AR, and constructing and superposing a virtual inspection area and a result by extracting model output and real-time parameters to realize visual real-time contact pin position and state display; s4, feedback of operators: the operator evaluates the quality of the pin according to the AR checking result and takes necessary measures; the system comprises a hardware layer, a data layer, a functional layer and an application layer.

Description

A method, device and system for intelligent inspection of the crimping quality of micro pins of electrical connectors

Technical Field

The present invention relates to the field of information technology, and in particular to a method, device and system for intelligently inspecting the crimping quality of a miniature pin of an electrical connector.

Background technique

Electrical connectors are core components of electrical circuit interconnection systems. Their function is to simplify electrical circuit connections of high-end equipment such as aircraft, ships, and high-speed railways, increase system connectivity, provide convenience for connections, and achieve decoupling of connection components. Pins are core components for interconnecting electrical connectors. The quality of their crimping is directly related to the power quality and operational stability of wires and cables. Therefore, it is extremely important to inspect the crimping quality of pins during pin production and electrical connector insertion to ensure that their quality meets the requirements.

Most of the pins are miniature parts, with a minimum diameter of only about 0.5mm. The area to be inspected is either in a miniature hole or on a narrow edge with a size less than 1mm. Each electrical connector requires a large number of pins, up to 128, which poses a great challenge to its quality inspection.

The wire pressing status inside the pin and the exposed length of the wire core are two key inspection points for the quality inspection of pin crimping. The wire pressing status inspection inside the pin is to check whether the cable core is fully inserted into the pin. It needs to be checked at a specific location and in the pin observation hole area with a diameter generally less than 2mm. The exposed length of the wire core conformity inspection is to check whether the exposed length of the wire core is within the specified length (generally less than 1mm). It is checked at the tail area of the pin. At present, both of these inspection points are highly dependent on manual inspection using a magnifying glass, which is very inconvenient to operate. The efficiency of large-scale pin inspection is very low, and it is easy to cause visual fatigue to the operators, and it is extremely easy to have problems such as wrong inspection and missed inspection.

At present, technologies such as deep learning, computer vision, and augmented reality are developing rapidly, and their applications in manufacturing quality inspection are becoming more and more extensive, with good application prospects. However, due to the small size of micro pins and the high difficulty of quality inspection, there is still a lack of intelligent inspection methods, devices, and systems for the crimping quality of micro pins of electrical connectors. The quality of pin crimping is still manually inspected using a magnifying glass, and real-time, automated, and intelligent auxiliary inspection methods and tools are urgently needed.

At present, the quality inspection of pin crimping mainly adopts the method of manual use of magnifying glasses, which is very inconvenient to operate and has a very low inspection efficiency. Large-scale inspection is easy to cause visual fatigue of operators, and problems such as wrong inspection and missed inspection occur frequently.

Summary of the invention

Therefore, the present invention proposes an intelligent inspection method, device and system for the crimping quality of micro-pins of electrical connectors, which can assist inspectors in performing real-time, intelligent and high-precision inspections of crimping quality points such as the wire pressing status inside the micro-pins and the compliance of the exposed length of the wire cores through methods such as deep learning, computer vision and augmented reality, thereby solving the problems of inconvenient operation, low efficiency, easy visual fatigue, easy misdetection and missed detection caused by manual inspection with a magnifying glass.

In order to achieve the above-mentioned object, the present invention provides an intelligent inspection device for the crimping quality of a micro pin of an electrical connector, the device comprising: a box body, a box cover and a camera device, the box body and the box cover are detachably connected, the box cover is fixedly connected to a ring light source on one side close to the box body, the camera device comprises a camera and a camera module protective shell, and the number of the camera device is at least one;

When there is only one camera device, a through hole for the pin to pass through is provided on one side of the box body, the camera device is provided on the side of the box cover away from the box body, a channel for the camera to pass through is provided at the center of the box cover, a fixing groove is provided around the channel, a fixing block is provided on the side of the camera module protective shell close to the camera, and the camera module protective shell is detachably connected to the box cover through the fixing block;

When there are multiple camera devices, the box cover is provided with a through hole for the pin to pass through, the box cover is provided with two, and the two box covers are arranged on two opposite surfaces of the box body, and the other surfaces of the box body except the surface to which the box cover is detachably connected are provided with channels for the camera to pass through, and fixing grooves are provided around the channels, and the camera module protective shell is provided with a fixing block on the side close to the camera, and the camera module protective shell is detachably connected to the box body through the fixing block.

Furthermore, the present invention also includes a method for intelligently inspecting the crimping quality of a micro-pin of an electrical connector, which is applied to the above-mentioned intelligent inspection device for the crimping quality of a micro-pin of an electrical connector, and the method includes:

S1. Offline modeling: The model is trained through deep learning, using high-quality, accurately annotated image datasets to detect and locate key parts of the pin in real time;

S2. Pin crimping quality inspection: Check the crimping status inside the pin and the exposed length of the wire core;

S3. AR display of inspection results: By extracting model output and real-time parameters, constructing and superimposing virtual inspection areas and results, an intuitive real-time display of pin position and status is achieved;

S4. Feedback: Evaluate the quality of the pins based on the AR inspection results and take action.

Preferably, step S1 specifically includes:

1) Prepare the data set: prepare some high-quality image data and build the data set;

2) Image partitioning: partitioning the high-resolution pin image in the high-quality image data, and using square areas to intercept the observation hole and tail images;

3) Image annotation: Accurately annotate the objects in each image to form a training data set;

4) Training and modeling: The labeled data set is trained through the deep learning target detection framework to obtain the target detection classifier, which is used to identify the pin observation hole and tail position in real time.

Preferably, step S2 specifically includes checking the needle inner wire pressing state and checking the wire core exposed length compliance;

The needle inner pressure line state inspection specifically includes:

Observation hole segmentation step: the sample image is segmented into multiple regions, and the position of the pin observation hole is identified by using the target detection classifier. By determining the position, the ROI region corresponding to the pin observation hole can be extracted, thereby realizing the segmentation of the observation hole.

Binarization process: Binarization and morphological operations are further performed on the extracted ROI area to distinguish the wire core from the inner wall of the observation hole. After the binarization process, the wire core appears white, while the inner wall of the observation hole appears black.

Boundary fitting method: The least square method is used to fit the boundary of the observation hole to an ellipse. The best fitting result is obtained by minimizing the residual square sum, ensuring that subsequent analysis is based on the precise hole boundary.

Parameter judgment mechanism: within the determined hole boundary, the proportion of the wire core area is calculated and a threshold is set. If the proportion of the wire core area is less than the threshold, it is determined that the cable core is not fully inserted into the crimping hole, that is, the crimping is not in place; if the proportion of the wire core area is greater than the threshold, it is determined that the cable core is fully inserted into the crimping hole, that is, the crimping is in place;

The wire core exposed length compliance check includes:

Tail segmentation step: performing image partitioning on the sample image so as to identify the position of the tail of the pin through the object detection classifier, and extracting the ROI area corresponding to the tail of the pin by determining the position, thereby realizing the segmentation of the tail of the pin;

Binarization process: further perform binarization and morphological operations on the extracted ROI area to remove interference information and improve the accuracy of subsequent analysis;

Contour extraction technology: The tail contour including the exposed wire core is extracted through Canny edge detection technology;

Method for determining the wire core area: The specific area of the exposed wire core is determined by measuring the change in profile width and angle;

Parameter judgment mechanism: The pixel distance of the exposed wire core length is converted into the actual distance through calibration. If the actual length of the exposed wire core is greater than the preset length threshold, it is judged as unqualified; if the actual length of the exposed wire core is less than or equal to the preset length threshold, it is judged as qualified.

Preferably, step S3 specifically includes:

Position information extraction method: extract real-time position information from the output of the deep learning model to accurately locate the coordinates of the pin observation hole and tail in the image;

Parameter information acquisition step: acquiring real-time parameter information of the inspection process, wherein the real-time parameter information includes inspection result data related to the needle inner wire pressing state and the exposed length of the wire core;

Inspection area frame selection technology: Based on the real-time position information, a virtual inspection area frame is constructed on the actual pin image to accurately mark the position of the pin observation hole and tail;

Inspection result presentation mechanism: Present inspection results based on real-time parameter information by means of virtual graphics, colors or marks to distinguish between qualified and unqualified status;

Virtual information overlay method: The inspection area frame and the inspection results are virtually overlaid on the real-time image, so that the virtual information is integrated with the actual pin image, thereby providing a more intuitive display of the inspection results.

Preferably, the parameter determination of the needle inner pressure line state inspection includes:

The computer determines the wire pressing condition according to the proportion of the wire core area;

The proportion of the wire core area is T0, the proportion of the first wire core area is T1, the proportion of the second wire core area is T2, the proportion of the third wire core area is T3, the proportion of the fourth wire core area is T4 and the proportion of the fifth wire core area is T5, and T1＜T2＜T3＜T4＜T5; the first-level wire pressing condition V1, the second-level wire pressing condition V2, the third-level wire pressing condition V3, the fourth-level wire pressing condition V4 and the fifth-level wire pressing condition V5 are preset, and V1＜V2＜V3＜V4＜V5;

Determine the wire pressing condition according to the relationship between the proportion T0 of the wire core area and the proportions of each preset wire core area;

When T0≤T1, the pressing line condition is determined to be the first-level pressing line condition V1;

When T1＜T0≤T2, the line pressing state is determined to be the secondary line pressing state V2;

When T2＜T0≤T3, the pressing line condition is determined to be the third-level pressing line condition V3;

When T3＜T0≤T4, the pressing line condition is determined to be the fourth-level pressing line condition V4;

When T4＜T0≤T5, the pressing line condition is determined to be the fifth-level pressing line condition V5;

When the crimping condition is V3, V4 or V5, it is determined that the crimping does not meet the technical requirements. The pins that are determined to be crimped do not meet the technical requirements are removed from the electrical connector, the condition of the wires and cables is checked, and then crimped again. Before re-crimping, it should be ensured that the conductor part of the wires and cables is not damaged, and the crimping tools and dies meet the requirements;

When the crimping condition is V1 or V2, it is determined that the crimping is firm and meets the technical requirements.

Preferably, the parameter judgment of the wire core exposed length compliance check includes:

The computer determines the actual length through the pixel distance of the exposed wire core length;

The pixel distance of the exposed core length is X0, and the pixel distance X1 of the first exposed core length, the pixel distance X2 of the second exposed core length, the pixel distance X3 of the third exposed core length, the pixel distance X4 of the fourth exposed core length, and the pixel distance X5 of the fifth exposed core length are preset, and X1＜X2＜X3＜X4＜X5; the first-level actual length N1, the second-level actual length N2, the third-level actual length N3, the fourth-level actual length N4, and the fifth-level actual length N5 are preset, and N1＜N2＜N3＜N4＜N5;

Determine the actual length according to the size relationship between the pixel distance X0 of the exposed wire core length and the pixel distances of each preset exposed wire core length;

When X0≤X1, the actual length is determined to be the first-level actual length N1;

When X1＜X0≤X2, the actual length is determined to be the secondary actual length N2;

When X2＜X0≤X3, the actual length is determined to be the third-level actual length N3;

When X3＜X0≤X4, the actual length is determined to be the fourth-level actual length N4;

When X4＜X0≤X5, the actual length is determined to be the fifth-level actual length N5;

When the actual length is N2, N3, N4 or N5, it is judged that the exposed part of the wire core is too long, and it needs to be trimmed to the specified length with a cutting tool. After trimming the wire core, it needs to be crimped again to ensure good electrical connection and mechanical strength;

When the actual length is V1, it is determined that the exposed portion of the wire core is not long, and there is no need to trim the wire core.

Furthermore, the present invention also includes an intelligent inspection system for the crimping quality of a micro-pin of an electrical connector, the system comprising a hardware layer, a data layer, a function layer and an application layer;

The hardware layer includes a computing device, an image acquisition device, a light source control device and a display device. The computing device is a computer for executing the inspection algorithm. The image acquisition device is a pin inspection box, which integrates a high-resolution short-focus camera and a 60-degree ring light source for image acquisition of actual pins. The light source control device is a light source controller for adjusting the brightness of the light source to ensure that the pins can obtain sufficient illumination. The display device is a touch screen for displaying real-time images and pin inspection results.

The data layer includes all the data in the system, including prior information, real-time position, parameter information and inspection result information; the prior information is divided into inspection process information, observation hole and tail image, and deep learning information; the real-time position and parameter information refer to the real-time two-dimensional position of the inspection area and the inspection parameter data within the area; the inspection result information refers to whether the inspection item is qualified;

The functional layer is a part that realizes intelligent inspection of pin crimping quality, including an inspection requirement design module, a pin inner crimping state inspection module, a wire core exposed length compliance inspection module and an inspection result AR display module. The inspection requirement design module is configured to provide prior information for subsequent functions; in the pin inner crimping state inspection module and the wire core exposed length compliance inspection module, deep learning information is first used to identify the target inspection area, and then real-time position information and parameter information are obtained, and finally the inspection result information is obtained by combining the inspection process information and parameter information;

The application layer represents the system workflow steps as follows:

S1. Offline modeling: The model is trained through deep learning and image datasets are used to detect and locate the key parts of the pins in real time;

S2. Pin crimping quality inspection: Check the crimping status of the pin and the exposed length of the wire core;

S3. AR display of inspection results: By extracting model output and real-time parameters, constructing and superimposing virtual inspection areas and results, an intuitive real-time display of pin position and status is achieved;

S4. Operator feedback: The operator evaluates the quality of the pin based on the AR inspection results and takes necessary measures.

Compared with the prior art, the method, device and system for intelligent inspection of the crimping quality of micro pins of an electrical connector provided by the embodiment of the present invention have the following beneficial effects:

(1) A method for checking the pressure line status of the needle based on deep learning and visual fusion was proposed. Based on the deep learning segmentation results, the method used the least squares method to perform ellipse fitting, successfully restored the boundary of the observation hole, and formed an accurate analysis area. The area ratio was calculated within the hole boundary, and a reliable judgment of the pressure line status of the observation hole was achieved. This method effectively solved the problem of hole boundary division and provided a new idea for checking the pressure line status of the needle.

(2) A method for checking the exposed length of wire cores based on deep learning and visual fusion is proposed. Based on the deep learning segmentation results, this method determines the precise position of the exposed wire core by measuring the change in contour width and angle. Combined with calibration and threshold judgment, accurate inspection of the exposed length of the wire core at the end of the pin is achieved. This method effectively solves the problem of judging the specific position of the wire core and provides a new idea for the accurate measurement of the exposed length of the wire core.

(3) An intelligent pin inspection device was designed. The device has two different types: a single-camera pin inspection box and a four-camera pin inspection box. By integrating multiple hardware components, including a high-resolution short-focus camera, a ring light source, and a reasonable box structure design, efficient inspection of the pins is achieved. This device not only takes into account the balance between cost and efficiency, but also fully considers the ease of operation and inspection accuracy, bringing a new hardware integration solution to the field of pin crimping quality inspection.

BRIEF DESCRIPTION OF THE DRAWINGS

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

FIG1 is a flow chart of an intelligent inspection method for crimping quality of a micro-pin of an electrical connector provided by an embodiment of the present invention;

2 is a flowchart of checking the internal pressure line state of a pin of an electrical connector micro pin crimping quality intelligent inspection method provided by an embodiment of the present invention;

3 is a flow chart of compliance inspection of exposed length of wire cores of an intelligent inspection method for crimping quality of micro pins of an electrical connector provided by an embodiment of the present invention;

4 is an AR display flow chart of the inspection results of an intelligent inspection method for crimping quality of micro pins of an electrical connector provided by an embodiment of the present invention;

5 is a structural diagram of a single-camera pin inspection box of an electrical connector micro pin crimping quality intelligent inspection device provided by an embodiment of the present invention;

6 is a structural diagram of a single-camera pin inspection box of an electrical connector micro pin crimping quality intelligent inspection device provided by an embodiment of the present invention;

7 is a structural diagram of a single-camera pin inspection box cover of an electrical connector micro pin crimping quality intelligent inspection device provided by an embodiment of the present invention;

8 is a structural diagram of a four-camera pin inspection box of an electrical connector micro pin crimping quality intelligent inspection device provided by an embodiment of the present invention;

9 is a structural diagram of a four-camera pin inspection box of an electrical connector micro pin crimping quality intelligent inspection device provided by an embodiment of the present invention;

10 is a structural diagram of a four-camera pin inspection box cover of an electrical connector micro pin crimping quality intelligent inspection device provided by an embodiment of the present invention;

11 is a structural diagram of a 60-degree annular light source of an intelligent inspection device for crimping quality of a micro-pin of an electrical connector provided by an embodiment of the present invention;

12 is a structural diagram of a high-resolution short-focus camera module of an intelligent inspection device for crimping quality of micro pins of an electrical connector provided by an embodiment of the present invention;

13 is a structural diagram of a camera module protective shell of an intelligent inspection device for crimping quality of micro pins of an electrical connector provided by an embodiment of the present invention;

14 is a hardware environment diagram of an intelligent inspection system for crimping quality of micro-pins of electrical connectors provided in an embodiment of the present invention;

FIG. 15 is a general structural framework diagram of an intelligent inspection system for the crimping quality of micro-pins of electrical connectors provided in an embodiment of the present invention.

In the figure, 1. box body; 2. box cover; 3. camera module protective shell; 4. camera; 5. ring light source; 6. computing device; 7. display device; 8. light source controller; 9. pin inspection box.

Detailed ways

The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

In the description of the present application, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, unless otherwise specified, "plurality" means two or more.

In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

As shown in FIGS. 1-4 , a preferred embodiment of an embodiment of the present invention provides an intelligent inspection method for the crimping quality of a micro-pin of an electrical connector, the method comprising:

S1. Offline modeling: The model is trained through deep learning, using high-quality, accurately annotated image datasets to detect and locate key parts of the pin in real time;

S2. Pin crimping quality inspection: Check the crimping status of the pin and the exposed length of the wire core;

S3. AR display of inspection results: By extracting model output and real-time parameters, constructing and superimposing virtual inspection areas and results, an intuitive real-time display of pin position and status is achieved;

S4. Feedback: Evaluate the quality of the pins based on the AR inspection results and take action.

It is understandable that the present invention uses a high-resolution short-focus camera to capture images, uses deep learning to further divide the target area, and proposes universal inspection methods for two types of pin crimping quality points: the pin inner wire pressure state and the wire core exposed length compliance, and uses AR technology to display the inspection results. Depending on the number of input images, the algorithm can be flexibly adjusted to adapt to both relatively low-cost single-camera inspection boxes and four-camera inspection boxes that meet the needs of more efficient inspections. The inspection method is divided into three parts: offline modeling, pin crimping quality inspection, and AR display of inspection results.

In this preferred embodiment, step S1 specifically includes:

1) Prepare the data set: prepare some high-quality image data and build the data set;

2) Image partitioning: partition the high-resolution pin image in the high-quality image data, and use square areas to capture the observation hole and tail images;

3) Image annotation: Accurately annotate the objects in each image to form a training data set;

4) Training and modeling: The labeled dataset is trained through the deep learning target inspection framework to obtain the target detection classifier, which is used to identify the pin observation hole and tail position in real time.

In this preferred embodiment, step S2 specifically includes checking the needle inner wire pressing state and checking the wire core exposed length compliance;

The needle inner pressure line state inspection specifically includes:

Image processing method: Segment the sample image into multiple regions, and use the target detection classifier to identify the position of the pin observation hole. By determining the position, the ROI region corresponding to the pin observation hole can be extracted, thereby realizing the segmentation of the observation hole.

Binarization processing step: further performing binarization processing and morphological operations on the extracted ROI area to distinguish the wire core from the inner wall of the observation hole. After the binarization processing, the wire core appears white, and the inner wall of the observation hole appears black;

Boundary fitting process: The least square method is used to fit the boundary of the observation hole to an ellipse, and the best fitting result is obtained by minimizing the residual square sum, so as to ensure that the subsequent analysis is based on the precise hole boundary;

Parameter judgment mechanism: within the determined hole boundary, the proportion of the wire core area is calculated and a threshold is set. If the proportion of the wire core area is less than the threshold, it is determined that the cable core is not fully inserted into the crimping hole, that is, the crimping is not in place; if the proportion of the wire core area is greater than the threshold, it is determined that the cable core is fully inserted into the crimping hole, that is, the crimping is in place;

The wire core exposed length compliance check includes:

Tail segmentation step: performing image partitioning on the sample image so as to identify the position of the tail of the pin through the object detection classifier, and extracting the ROI area corresponding to the tail of the pin by determining the position, thereby realizing the segmentation of the tail of the pin;

Binarization process: further perform binarization and morphological operations on the extracted ROI area to remove interference information and improve the accuracy of subsequent analysis;

Contour extraction technology: The tail contour including the exposed wire core is extracted through Canny edge detection technology;

Method for determining the wire core area: The specific area of the exposed wire core is determined by measuring the change in outline width and angle;

Parameter judgment mechanism: The pixel distance of the exposed wire core length is converted into the actual distance through calibration. If the actual length of the exposed wire core is greater than the preset length threshold, it is judged as unqualified; if the actual length of the exposed wire core is less than or equal to the preset length threshold, it is judged as qualified.

In this preferred embodiment, step S3 specifically includes:

Position information extraction method: extract real-time position information from the output of the deep learning model to accurately locate the coordinates of the pin observation hole and tail in the image;

Parameter information acquisition step: acquiring real-time parameter information of the inspection process, wherein the real-time parameter information includes inspection result data related to the needle inner wire pressing state and the exposed length of the wire core;

Inspection area frame selection technology: Based on the real-time position information, a virtual inspection area frame is constructed on the actual pin image to accurately mark the position of the pin observation hole and tail;

Inspection result presentation mechanism: Present inspection results based on real-time parameter information by means of virtual graphics, colors or marks to distinguish between qualified and unqualified status;

Virtual information overlay method: The inspection area frame and the inspection results are virtually overlaid on the real-time image, so that the virtual information is integrated with the actual pin image, thereby providing a more intuitive display of the inspection results.

In this preferred embodiment, the parameter determination of the needle inner pressure line state inspection includes:

The computer determines the wire pressing condition according to the proportion of the wire core area;

The proportion of the wire core area is T0, the proportion of the first wire core area is T1, the proportion of the second wire core area is T2, the proportion of the third wire core area is T3, the proportion of the fourth wire core area is T4 and the proportion of the fifth wire core area is T5, and T1＜T2＜T3＜T4＜T5; the first-level wire pressing condition V1, the second-level wire pressing condition V2, the third-level wire pressing condition V3, the fourth-level wire pressing condition V4 and the fifth-level wire pressing condition V5 are preset, and V1＜V2＜V3＜V4＜V5;

Determine the wire pressing condition according to the relationship between the proportion T0 of the wire core area and the proportions of each preset wire core area;

When T0≤T1, the pressing line condition is determined to be the first-level pressing line condition V1;

When T1＜T0≤T2, the line pressing state is determined to be the secondary line pressing state V2;

When T2＜T0≤T3, the pressing line condition is determined to be the third-level pressing line condition V3;

When T3＜T0≤T4, the pressing line condition is determined to be the fourth-level pressing line condition V4;

When T4＜T0≤T5, the pressing line condition is determined to be the fifth-level pressing line condition V5;

When the crimping condition is V3, V4 or V5, it is determined that the crimping does not meet the technical requirements. The pins that are determined to be crimped do not meet the technical requirements are removed from the electrical connector, the condition of the wires and cables is checked, and then crimped again. Before re-crimping, it should be ensured that the conductor part of the wires and cables is not damaged, and the crimping tools and dies meet the requirements;

When the crimping condition is V1 or V2, it is determined that the crimping is firm and meets the technical requirements.

In this preferred embodiment, the parameter judgment of the wire core exposed length compliance check includes:

The computer determines the actual length through the pixel distance of the exposed wire core length;

The pixel distance of the exposed core length is X0, and the pixel distance X1 of the first exposed core length, the pixel distance X2 of the second exposed core length, the pixel distance X3 of the third exposed core length, the pixel distance X4 of the fourth exposed core length, and the pixel distance X5 of the fifth exposed core length are preset, and X1＜X2＜X3＜X4＜X5; the first-level actual length N1, the second-level actual length N2, the third-level actual length N3, the fourth-level actual length N4, and the fifth-level actual length N5 are preset, and N1＜N2＜N3＜N4＜N5;

Determine the actual length according to the size relationship between the pixel distance X0 of the exposed wire core length and the pixel distances of each preset exposed wire core length;

When X0≤X1, the actual length is determined to be the first-level actual length N1;

When X1＜X0≤X2, the actual length is determined to be the secondary actual length N2;

When X2＜X0≤X3, the actual length is determined to be the third-level actual length N3;

When X3＜X0≤X4, the actual length is determined to be the fourth-level actual length N4;

When X4＜X0≤X5, the actual length is determined to be the fifth-level actual length N5;

When the actual length is N2, N3, N4 or N5, it is judged that the exposed part of the wire core is too long, and it needs to be trimmed to the specified length with a cutting tool. After trimming the wire core, it needs to be crimped again to ensure good electrical connection and mechanical strength;

When the actual length is V1, it is determined that the exposed portion of the wire core is not long, and there is no need to trim the wire core.

In order to cooperate with the above-mentioned inspection algorithm, the present invention designs an inspection box that integrates multiple hardware. The inspection boxes are divided into two types, one is a single-camera pin inspection box 9, and the other is a four-camera pin inspection box 9. The two types have their own characteristics to meet different needs. The single-camera pin inspection box 9 has certain advantages in terms of cost because only one camera is needed, but the single camera cannot capture every side of the pin at the same time. The operator needs to manually adjust the direction of the pin during the inspection, which will have a slight impact on efficiency. The four-camera pin inspection box 9 is fully automated. The four cameras can capture pin images from four directions. The operator does not need to manually adjust the direction of the pin. The inspection efficiency is higher, and the cost is relatively increased. Enterprises can choose a suitable inspection box according to their own characteristics and needs.

The specific design of the pin inspection box 9 is as follows:

1. Box structure design: In order to ensure stability and easy manufacturing and installation, a square structure design is selected.

2. Socket design: Sockets are designed on the two opposite sides of the inspection box to facilitate the insertion of wires for inspection. For a single-phase inspection box, holes are opened in the center of the two side walls of the box body; for a four-phase inspection box, holes are opened in the center of the two box covers. In order to limit the insertion direction of the pin, a glass tube is set between the two sides of the jack to ensure that the pin can only be inserted in the specified direction to avoid mis-inspection due to bending of the wire with the pin.

3. Light source selection: The selection of light source is crucial for pin inspection, because uniform lighting can reduce shadows and improve image quality. In order to improve the clarity and quality of the image, a 60-degree ring light source is used to ensure uniform lighting of the entire socket area. For a single-camera inspection box, a 60-degree ring light source is installed on the inside of the box cover; and for a four-camera inspection box, two 60-degree ring light sources are installed on both sides of the box body where the socket is located.

4. Camera selection: The pin size is small, and ordinary cameras are difficult to capture clear and accurate pin images. Choosing a suitable camera is crucial. To ensure that the camera can capture high-quality images at a close distance, a high-resolution short-focus camera is used. For a single-camera inspection box, a camera module is integrated on the box cover; for a four-camera inspection box, four camera modules are integrated on the four sides without jacks.

In general, the design of the pin inspection box 9 takes into account multiple aspects such as cost, ease of operation, and inspection accuracy. Through reasonable hardware integration and structural design, the efficiency and reliability of pin inspection are ensured. The selection of the square structure ensures the stability of the box body. The jack design limits the insertion direction of the pin by setting a glass tube to reduce the possibility of mis-inspection. The use of a ring light source and a high-resolution short-focus camera improves the clarity and quality of the image and ensures the accuracy of the pin inspection. Such a design scheme ensures the efficiency and reliability of the pin inspection while meeting different needs.

Please refer to Figures 5 to 13 below to describe in detail the structure of the pin inspection box 9:

An intelligent inspection device for the crimping quality of a micro pin of an electrical connector, the device comprising: a box body 1, a box cover 2 and a camera device, the box body 1 and the box cover 2 are detachably connected, a ring light source 5 is fixedly connected to the box cover 2 on a side close to the box body 1, the camera device comprises a camera 4 and a camera module protective shell 3, and the number of the camera device is at least one;

When there is only one camera device, a through hole for the pin to pass through is provided on one side of the box body 1, the camera device is provided on the side of the box cover 2 away from the box body 1, a channel for the camera 4 to pass through is provided in the center of the box cover 2, a fixing groove is provided around the channel, a fixing block is provided on the side of the camera module protective shell 3 close to the camera 4, and the camera module protective shell 3 is detachably connected to the box cover 2 through the fixing block;

When there are multiple camera devices, the box cover 2 is provided with a through hole for the pin to pass through, and two box covers 2 are provided, and the two box covers 2 are arranged on two opposite surfaces of the box body 1, and the other surfaces of the box body 1 except the surface to which the box cover 2 is detachably connected are provided with channels for the camera 4 to pass through, and fixing grooves are provided around the channels. The camera module protective shell 3 is provided with a fixing block on the side close to the camera 4, and the camera module protective shell 3 is detachably connected to the box body 1 through the fixing block.

Based on the above inspection method and inspection device, the present invention has developed an intelligent inspection system for the crimping quality of electrical connector micro pins, as shown in FIGS. 14 and 15 , wherein the system includes a hardware layer, a data layer, a function layer and an application layer;

The hardware layer includes a computing device 6, an image acquisition device, a light source control device and a display device 7. The computing device 6 is a computer for executing an inspection algorithm. The image acquisition device is a pin inspection box 9. The pin inspection box 9 integrates a high-resolution short-focus camera and a 60-degree ring light source for image acquisition of actual pins. The light source control device is a light source controller 8 for adjusting the brightness of the light source to ensure that the pins can obtain sufficient illumination. The display device 7 is a touch screen for displaying real-time images and pin inspection results.

The data layer includes all the data in the system, including prior information, real-time position, parameter information and inspection result information; the prior information is divided into inspection process information, observation hole and tail image, and deep learning information; the real-time position and parameter information refer to the real-time two-dimensional position of the inspection area and the inspection parameter data within the area; the inspection result information refers to whether the inspection item is qualified;

The functional layer is a part that realizes intelligent inspection of pin crimping quality, including an inspection requirement design module, a pin inner crimping state inspection module, a wire core exposed length compliance inspection module and an inspection result AR display module. The inspection requirement design module is configured to provide prior information for subsequent functions; in the pin inner crimping state inspection module and the wire core exposed length compliance inspection module, deep learning information is first used to identify the target inspection area, and then real-time position information and parameter information are obtained, and finally the inspection result information is obtained by combining the inspection process information and parameter information;

The application layer represents the system workflow steps as follows:

S1. Offline modeling: The model is trained through deep learning and image datasets are used to detect and locate the key parts of the pins in real time;

S2. Pin crimping quality inspection: Check the crimping status of the pin and the exposed length of the wire core;

S3. AR display of inspection results: By extracting model output and real-time parameters, constructing and superimposing virtual inspection areas and results, an intuitive real-time display of pin position and status is achieved;

S4. Operator feedback: The operator evaluates the quality of the pin based on the AR inspection results and takes necessary measures.

In summary, the embodiment of the present invention provides an intelligent inspection method, device and system for the crimping quality of a micro-pin of an electrical connector. Compared with the prior art, it proposes a method for inspecting the inner pressure line state of a pin based on deep learning and visual fusion. Based on the deep learning segmentation results, the method adopts the least squares method to perform ellipse fitting, successfully restores the boundary of the observation hole, and forms an accurate analysis area. Thus, the area ratio is calculated within the hole boundary, and a reliable judgment of the inner pressure line state of the observation hole is realized. This method effectively solves the problem of hole boundary division and provides a new idea for the inspection of the inner pressure line state of the pin. A method for inspecting the exposed length of a wire core based on deep learning and visual fusion is also proposed. Based on the deep learning segmentation results, the method determines the precise position of the exposed wire core by measuring the change in the contour width and the change in the angle. Combined with calibration and threshold judgment, the exposed length of the wire core at the tail of the pin is accurately inspected. This method effectively solves the problem of judging the specific position of the wire core and provides a new idea for the accurate measurement of the exposed length of the wire core. A pin intelligent inspection device is also designed. The device has two different types, namely a single-phase pin inspection box 9 and a four-phase pin inspection box 9. By integrating multiple hardware components, including a high-resolution short-focus camera, a ring light source, and a reasonable box structure design, efficient inspection of the pins is achieved. This device not only takes into account the balance between cost and efficiency, but also fully considers the ease of operation and inspection accuracy, bringing a new hardware integration solution to the field of pin crimping quality inspection.

Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

The present application is described with reference to the flowcharts and/or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and/or box in the flowchart and/or block diagram, as well as the combination of the processes and/or boxes in the flowchart and/or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions specified in one or more processes in the flowchart and/or one or more boxes in the block diagram.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device that implements the functions specified in one or more processes in the flowchart and/or one or more boxes in the block diagram.

These computer program instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and/or one or more boxes in the block diagram.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims (8)

1. An intelligent inspection device for the press-fit quality of a miniature pin of an electric connector, which is characterized by comprising: the camera comprises a camera and a camera module protective shell, wherein at least one camera is arranged in the camera module protective shell;

when only one camera device exists, a through hole for a pin to pass through is formed in one side of the box body, the camera device is arranged on one side of the box cover, which is far away from the box body, a channel for a camera to pass through is formed in the center of the box cover, a fixing groove is formed in the periphery of the channel, a fixing block is arranged on one side, which is close to the camera, of the camera module protection shell, and the camera module protection shell is detachably connected with the box cover through the fixing block;

When the camera device has a plurality of camera devices, through holes through which the contact pins can pass are formed in the box cover, two box covers are arranged on two opposite surfaces of the box body, channels through which the camera can pass are formed in other surfaces of the box body except the surfaces of the box cover which are detachably connected, fixing grooves are formed in the periphery of the channels, a fixing block is arranged on one side of the camera module protective shell, which is close to the camera, of the camera module protective shell, and the camera module protective shell is detachably connected with the box body through the fixing block.

2. An intelligent inspection method for the press-connection quality of a micro-pin of an electric connector, which is characterized in that the method is applied to the intelligent inspection device for the press-connection quality of the micro-pin of the electric connector in claim 1, and comprises the following steps:

s1, offline modeling: detecting and positioning key parts of the contact pin in real time by using an image dataset through a deep learning training model;

S2, checking the press-connection quality of the pins: performing in-needle pressure line state inspection and wire core exposed length compliance inspection;

S3, displaying an inspection result AR: by extracting model output and real-time parameters, a virtual inspection area and a result are constructed and overlapped, and visual real-time contact pin position and state display is realized;

s4, feedback: and evaluating the quality of the pin according to the AR checking result, and taking measures.

3. The intelligent inspection method for the press-fit quality of the micro pin of the electrical connector according to claim 2, wherein the step S1 specifically comprises:

1) Preparing a data set: preparing a plurality of high-quality image data and constructing a data set;

2) Image partitioning: partitioning a high-resolution pin image in the high-quality image data, and intercepting an observation hole and a tail image by using a square area;

3) Image marking: accurately labeling targets in each picture to form a training data set;

4) Training and modeling: training the marked data set through a deep learning target detection framework to obtain a target detection classifier for identifying the positions of the contact pin observation holes and the tail parts in real time.

4. The intelligent inspection method for the compression quality of the micro-pin of the electric connector according to claim 2, wherein the step S2 specifically comprises an in-pin compression state inspection and a wire core exposed length compliance inspection;

the in-needle pressure state check specifically includes:

The image processing method comprises the following steps: dividing a sample image into a plurality of areas, identifying the positions of the contact pin observation holes by using a target detection classifier, and extracting the ROI areas corresponding to the contact pin observation holes by determining the positions so as to realize the division of the observation holes;

Binarization processing step: further performing binarization processing and morphological operation on the extracted ROI region to distinguish the wire core from the inner wall of the observation hole, wherein the wire core is white and the inner wall of the observation hole is black after the binarization processing;

Boundary fitting process: elliptical fitting is carried out on the boundary of the observation hole by adopting a least square method, and the best fitting result is obtained by minimizing the sum of squares of residual errors, so that the follow-up analysis is ensured to be carried out based on the accurate hole boundary;

parameter judgment mechanism: in the determined hole boundary, calculating the ratio of the area of the wire core, setting a threshold value, and if the ratio of the area of the wire core is smaller than the threshold value, judging that the wire core of the cable is not completely inserted into the crimping hole, namely the wire pressing is not in place; if the area ratio of the wire core is larger than the threshold value, judging that the cable wire core is completely inserted into the crimping hole, namely, pressing the wire into place;

the core bare length compliance check includes:

Tail segmentation: carrying out image partitioning on the sample image, identifying the position of the tail of the pin through a target detection classifier, and extracting the ROI area corresponding to the tail of the pin through determining the position so as to realize the segmentation of the tail of the pin;

binarization processing process: further performing binarization processing and morphological operation on the extracted ROI area to remove interference information;

contour extraction technology: extracting tail contours comprising bare wire cores by a Canny edge detection technology;

The wire core area determining method comprises the following steps: determining a specific area of the exposed wire core by measuring and calculating the change of the width and the change of the angle of the profile;

Parameter judgment mechanism: the pixel distance of the length of the exposed wire core is converted into the actual distance through calibration, and if the actual length of the exposed wire core is larger than a preset length threshold value, the exposed wire core is judged to be unqualified; and if the actual length of the exposed wire core is smaller than or equal to the preset length threshold value, judging that the wire core is qualified.

5. The intelligent inspection method for the press-fit quality of the micro pin of the electrical connector according to claim 2, wherein the step S3 specifically comprises:

the position information extraction method comprises the following steps: extracting real-time position information from the output of the deep learning model to accurately position coordinates of the pin observation hole and the tail in the image;

parameter information acquisition: acquiring real-time parameter information of an inspection process, wherein the real-time parameter information comprises inspection result data of an in-needle line pressing state and a wire core exposed length;

Inspection area framing technique: constructing a virtual checking area frame on the actual contact pin image according to the real-time position information so as to accurately mark the positions of the contact pin observation hole and the tail part;

Inspection result presentation mechanism: presenting the checking result according to the real-time parameter information in a virtual graph, color or mark mode so as to distinguish the qualified state and the unqualified state;

The virtual information superposition method comprises the following steps: and superposing the inspection area frame and the inspection result on the real-time image in a virtual mode, so that virtual information and an actual contact pin image are integrated, and more visual inspection result display is provided.

6. The intelligent inspection method for the press-fit quality of a micro-pin of an electrical connector according to claim 4, wherein the parameter judgment for the in-pin press-fit state inspection comprises:

The computer determines the line pressing condition through the ratio of the area of the line core;

The ratio of the wire core areas is T0, the ratio T1 of the first wire core area, the ratio T2 of the second wire core area, the ratio T3 of the third wire core area, the ratio T4 of the fourth wire core area and the ratio T5 of the fifth wire core area are preset, and T1 is more than T2 and less than T3 and less than T5; presetting a primary line pressing condition V1, a secondary line pressing condition V2, a tertiary line pressing condition V3, a quaternary line pressing condition V4 and a fifth line pressing condition V5, wherein V1 is more than V2 and less than V3 and less than V4 and less than V5;

determining a wire pressing condition according to the relation between the occupied ratio T0 of the wire core area and the occupied ratio of each preset wire core area;

when T0 is less than or equal to T1, determining the line pressing condition as a primary line pressing condition V1;

When T1 is more than or equal to T0 and less than or equal to T2, determining the line pressing condition as a secondary line pressing condition V2;

When T2 is more than or equal to T0 and less than or equal to T3, determining the line pressing condition as a three-level line pressing condition V3;

when T3 is more than or equal to T0 and less than or equal to T4, determining the line pressing condition as a four-stage line pressing condition V4;

When T4 is more than or equal to T0 and less than or equal to T5, determining the line pressing condition as a five-stage line pressing condition V5;

When the line pressing condition is V3, V4 or V5, judging that the crimping is not in accordance with the technical requirement, judging that the crimping is not in accordance with the contact pin of the technical requirement, taking out the contact pin from the electric connector, checking the state of the electric wire and the electric cable, and then re-crimping, wherein before re-crimping, the conductor part of the electric wire and the electric cable is ensured to be free from damage, and the crimping tool and the die are in accordance with the requirement;

and when the line pressing condition is V1 or V2, the crimping is judged to be firm, and the technical requirement is met.

7. The intelligent inspection method for the compression quality of a micro-pin of an electrical connector according to claim 4, wherein the parameter judgment for the compliance inspection of the exposed length of the wire core comprises:

the computer determines the actual length through the pixel distance of the length of the exposed wire core;

The pixel distance of the length of the exposed wire core is X0, the pixel distance X1 of the length of the first exposed wire core, the pixel distance X2 of the length of the second exposed wire core, the pixel distance X3 of the length of the third exposed wire core, the pixel distance X4 of the length of the fourth exposed wire core and the pixel distance X5 of the length of the fifth exposed wire core are preset, and X1 is more than X2 and less than X3 and less than X5; presetting a primary actual length N1, a secondary actual length N2, a tertiary actual length N3, a quaternary actual length N4 and a penta actual length N5, wherein N1 is more than N2 and less than N3 and less than N4 and less than N5;

Determining the actual length according to the relation between the pixel distance X0 of the length of the exposed wire core and the pixel distance of each preset length of the exposed wire core;

When X0 is less than or equal to X1, determining the actual length as a first-stage actual length N1;

When X1 is more than X0 and less than or equal to X2, determining the actual length as a secondary actual length N2;

when X2 is more than X0 and less than or equal to X3, determining the actual length as three-stage actual length N3;

When X3 is more than X0 and less than or equal to X4, determining the actual length as a four-stage actual length N4;

when X4 is more than X0 and less than or equal to X5, determining the actual length to be five-stage actual length N5;

when the actual length is N2, N3, N4 or N5, judging that the exposed part of the wire core is too long, trimming the wire core to a specified length by using a cutting tool, and after trimming the wire core, crimping the wire core again to ensure good electrical connection and mechanical strength;

when the actual length is V1, judging that the exposed part of the wire core is not long, and trimming the wire core is not needed.

8. An intelligent checking system for the press-connection quality of a miniature contact pin of an electric connector is characterized by comprising a hardware layer, a data layer, a functional layer and an application layer;

The hardware layer comprises a computing device, an image acquisition device, a light source control device and a display device, wherein the computing device is a computer and is used for executing an inspection algorithm, the image acquisition device is a pin inspection box, the pin inspection box is integrated with a high-resolution short-focus camera and a 60-degree annular light source and is used for image acquisition of an actual pin, the light source control device is a light source controller and is used for adjusting the brightness of the light source so as to ensure that the pin can obtain sufficient illumination, and the display device is a touch display screen and is used for displaying a real-time image and a pin inspection result;

The data layer comprises all data in the system, including prior information, real-time position, parameter information and inspection result information; the prior information is divided into inspection process information, observation holes, tail images and deep learning information, and the real-time position and parameter information refer to real-time two-dimensional position of an inspection area and inspection parameter data in the area; the checking result information refers to whether the checking item is qualified or not;

the functional layer is a part for realizing intelligent inspection of the pin press-connection quality and comprises an inspection requirement design module, an in-needle pressure line state inspection module, a wire core exposed length compliance inspection module and an inspection result AR display module, wherein the inspection requirement design module is configured to provide prior information for subsequent functions; in the in-needle pressure linear state checking module and the wire core exposure length compliance checking module, firstly, a target checking area is identified by using deep learning information, then real-time position information and parameter information are obtained, and finally checking process information and parameter information are combined to obtain checking result information;

The application layer representation system workflow steps are as follows:

s1, offline modeling: detecting and positioning key parts of the contact pin in real time by using an image dataset through a deep learning training model;

S2, checking the press-connection quality of the pins: performing in-needle pressure line state inspection and wire core exposed length compliance inspection;

S3, displaying an inspection result AR: by extracting model output and real-time parameters, a virtual inspection area and a result are constructed and overlapped, and visual real-time contact pin position and state display is realized;

s4, feedback of operators: the operator evaluates the quality of the pin according to the AR inspection result and takes necessary measures.