Self-adaptive full-view-angle PCB image acquisition and two-dimensional code recognition filing system and method

Abstract

The invention discloses a self-adaptive full-view-angle PCB image acquisition and two-dimensional code recognition filing system and method. The system comprises a hardware end and a software end. The hardware end comprises a visual system module, a light source controller, a motion lifting module, a motion controller and an edge computer; the software end comprises an image acquisition module, a detection and identification module, a motion control module and a data decision module; the automatic process of PCB fine two-dimensional code detection and identification, model and number acquisition, multi-level folder creation, adaptive height full-angle image acquisition and archiving according to the two-dimensional code identification result is realized. Manual code scanning is not needed, and a multi-level folder path is created according to a code recognition result, so that the time and labor cost is reduced, the consistency of the result is ensured, and the misoperation condition caused by fatigue due to repeated manual work is avoided; the problems of low image acquisition and two-dimensional code recognition efficiency, limited visual angle, insufficient accuracy and the like in the traditional PCB manufacturing process are effectively solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of PCB image acquisition and two-dimensional code recognition, and in particular relates to an adaptive full-view PCB image acquisition and two-dimensional code recognition archiving system and method. Background Art

[0002] In existing PCB (printed circuit board) manufacturing processes, image acquisition and QR code recognition typically rely on traditional methods, such as manual operation or fixed camera equipment. However, these methods have some obvious shortcomings that limit production efficiency, accuracy, and automation levels.

[0003] Traditional manual image acquisition methods are inefficient, requiring significant manpower and time. Operator fatigue and experience can easily affect image acquisition, leading to inaccurate images or missed critical details. This reduces production efficiency and increases the risk of human error. Traditional fixed camera equipment also has limitations in image acquisition and QR code recognition. They typically capture images from a limited number of angles or specific viewing angles, failing to fully capture all areas and surface details of the PCB. This can lead to certain defects or QR code information being overlooked or not being recognized in a timely manner. Furthermore, these traditional methods lack real-time feedback and automatic adjustments. They lack timely image processing and recognition results, and cannot seamlessly integrate and collaborate with the entire manufacturing process. This prevents manufacturers from promptly identifying and correcting problems during production, impacting product quality and manufacturing efficiency. Summary of the Invention

[0004] In response to the problems existing in the existing technology, the present invention provides an adaptive full-view PCB image acquisition and QR code recognition and archiving system and method, designs a lifting platform, and utilizes a two-layer deep learning model and a high-precision recognition algorithm to improve the comprehensiveness, accuracy and stability of detection.

[0005] The technical solution of the present invention is achieved as follows:

[0006] Adaptive full-view PCB image acquisition and QR code recognition archiving system, including hardware and software;

[0007] The hardware side includes a visual system module, a light source controller, a motion lifting module, a motion controller and an edge computer; the software side includes an image acquisition module, a detection and recognition module, a motion control module and a data decision module;

[0008] The motion lifting module is provided with a lifting structure; the visual system module is provided with an industrial camera and a bar light source, and PCB images and QR code images are acquired through the motion lifting module and the visual system module; the light source controller is used to control the bar light source; the motion controller is used to drive the stepper motor to rotate and convert the motion mode through the edge computer;

[0009] The motion control module controls the motion of the lifting structure in the motion lifting module through a motion controller; the image acquisition module performs image acquisition and parameter configuration by controlling the visual system module; the detection and recognition module deploys a two-layer deep learning model in the software architecture of the edge computer, detects each QR code and stores information, and optionally creates multi-level folders and updates the full-view image save path.

[0010] The data decision module is used to store images from various perspectives; the edge computer makes decisions and manages data by controlling each module.

[0011] Furthermore, the lifting structure of the motion lifting module includes a ball screw guide rail, a stepping motor, a guide rail slider, a screw slider, a slider connecting plate, a right-angle adapter and a platform plate;

[0012] The stepper motor is arranged on the upper part of the ball screw guide rail to drive the ball screw guide rail. The guide rail slider is arranged on the ball screw guide rail. The screw slider is installed on the ball screw guide rail through the guide rail slider and moves up and down along the ball screw guide rail. The screw slider is connected to two right-angle adapters through a slider connecting plate. The right-angle adapter is connected to the platform plate. The platform plate reciprocates up and down with the screw slider.

[0013] Furthermore, the visual system module includes an industrial camera, a camera lens, a camera gripper, a bracket pole, a light source bracket and a bar light source;

[0014] The camera lens is located on the industrial camera and faces the platform board; the industrial camera is set on the camera clamp, the camera clamp is connected to the bracket pole, and the camera clamp is adjustable; a light source bracket is also provided on the bracket pole, the light source bracket is located below the camera clamp, and the light source bracket is adjustable, and two strip light sources are provided on the light source bracket, which uniformly illuminate the platform board from both sides.

[0015] Furthermore, the ball screw guide rail is arranged parallel to the bracket upright, the visual system module and the motion lifting module are located on the same board, and a light source controller, a motion controller and an edge computer are provided below the visual system module and the motion lifting module.

[0016] Furthermore, the light source controller drives and controls two bar light sources, adjusts the light source power through digital signals to control the brightness of the light source; the motion controller drives and controls the rotation of the stepper motor, receives the pulse level signal from the edge computer, drives the rotation of the stepper motor, and converts the rotational motion into the linear reciprocating motion of the screw slider through the ball screw guide.

[0017] Furthermore, the image acquisition module is used to control the shooting process of the industrial camera from the visual system module and read the real-time image stream of the industrial camera; specifically, it includes:

[0018] Initialize the industrial camera, configure the parameters of the industrial camera for QR code detection, trigger image acquisition according to the instructions sent by the edge computer, digitize the image data and transmit it to the edge computer for subsequent detection and recognition operations;

[0019] After detection and recognition are completed, the industrial camera parameters are switched to the configuration corresponding to full-view PCB image acquisition, and the image stream is transmitted to the edge computer for subsequent batch archiving and storage.

[0020] Furthermore, the detection and identification module detects the QR code object based on the collected PCB image and identifies the information in the QR code object;

[0021] The edge computer's software architecture deploys two layers of deep learning models. The first layer detects and segments small QR code objects within a large field of view. The second layer identifies the content of the separated QR code objects and obtains the string content information of the QR code objects.

[0022] The detection and identification module identifies and detects one or more two-dimensional code objects, and stores the detection and character string content information of each two-dimensional code object.

[0023] Furthermore, the legitimacy of the string content information of the QR code object is checked against the preset rules, and the string content information is processed through regularized matching to obtain the model and number of the product corresponding to the PCB image. Based on the string content information, multi-level folders are created and the full-view image save path is updated.

[0024] Furthermore, the motion control module receives the level pulse information converted by the edge computer based on the lifting distance information, and controls the lifting structure of the motion lifting module to perform linear reciprocating motion through the motion controller and the stepping motor;

[0025] The data decision module is used to store the images and names of each perspective corresponding to the PCB image, and after checking the legitimacy of the image saving path, batch save the full-perspective images locally, and then record each process and visualize the results.

[0026] The adaptive full-view PCB image acquisition and QR code recognition and archiving method applies the adaptive full-view PCB image acquisition and QR code recognition and archiving system described in any one of the above. The specific method is as follows:

[0027] The user places the PCB to be processed on the platform board, which is then raised or lowered to the height required for QR code detection. The industrial camera is configured for detection and recognition based on the current parameter information. The detection and recognition module detects and recognizes the QR code, detects and recognizes all QR code objects, and annotates and visualizes the results of multiple QR code detections.

[0028] The user selects a QR code object for creating a new multi-level folder for archiving. Based on the model and number information of the selected QR code object, a multi-level folder is created, a save path for the updated full-view image is generated, and the current parameter information of the industrial camera is switched to the full-view acquisition configuration. The platform plate is raised or lowered to the corresponding height according to the size of the PCB. The software end collects and names images from different perspectives, and then triggers batch saving of images to complete the archiving task of the PCB based on model and number.

[0029] Compared with the prior art, the present invention achieves the following beneficial effects:

[0030] The present invention provides an adaptive full-view PCB image acquisition and QR code recognition and archiving system and method, and designs an automated detection and archiving hardware system including a visual system module, a light source controller, a motion lifting module, a motion controller and an edge computer. The motion lifting module is adapted to the QR code recognition and image acquisition of PCBs of different three-dimensional sizes, including coping with the problems of different lengths, widths and heights, and different fields of view and depths of field at different viewing angles. The visual system module ensures that even small QR code objects can obtain sufficient pixel points, thereby obtaining high-quality images, providing stability and efficiency for subsequent detection and recognition, while also ensuring high-quality acquisition of full-view archived images.

[0031] An automated detection and archiving software system was also designed, which includes an image acquisition module, a detection and recognition module, a motion control module, and a data decision module. The image acquisition module is used to control the shooting process of the industrial camera and read the real-time image stream of the industrial camera. The detection and recognition module is used to detect and recognize small QR codes on PCBs, and regularize the matching to obtain product models and numbers, automatically create multi-level folders, and set archiving paths. The motion control module’s action execution instructions control the platform’s lifting and lowering to adapt to PCB objects of different heights and sizes, thereby realizing the automation of the entire process.

[0032] The adaptive full-view PCB image acquisition and QR code recognition and archiving system provided by this invention implements an automated process for detecting and recognizing small QR codes on PCBs, acquiring model and serial numbers, creating multi-level folders, and adaptively capturing highly full-view images and archiving them based on the QR code recognition results. This eliminates the need for manual code scanning and creating multi-level folders based on the recognition results, reducing time and labor costs, ensuring consistent results, and avoiding operational errors caused by repetitive labor and fatigue. The system can capture full-view images of a PCB and automatically archive them based on the product model and serial number contained in the QR code within 5 seconds, significantly improving efficiency compared to traditional manual processing.

[0033] A large number of images of PCBs containing QR codes are collected and a deep learning model for QR code object detection is trained based on a mature network framework. A second-layer pre-trained deep learning model for identifying QR code content is built. By combining the two-layer model, a dynamic ROI effect is achieved, further enabling the detection and content decoding of small-sized QR codes in a large field of view, reducing false detection and missed detection rates, and improving detection efficiency.

[0034] This invention effectively addresses the problems of inefficient image acquisition and QR code recognition, limited viewing angles, and insufficient accuracy in traditional PCB manufacturing. This system provides the PCB manufacturing industry with an efficient and accurate solution for full-view image acquisition and batch archiving, significantly improving product quality control and production efficiency, and driving the manufacturing industry towards intelligent and efficient manufacturing. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a system composition diagram of the adaptive full-view PCB image acquisition and QR code recognition and archiving system provided by an embodiment of the present invention;

[0036] Figure 2 2. It is a schematic structural diagram of a motion lifting module of an adaptive full-view PCB image acquisition and QR code recognition and archiving system provided by an embodiment of the present invention;

[0037] Figure 3 This is a schematic diagram of the structure of the visual system module of the adaptive full-view PCB image acquisition and QR code recognition and archiving system provided by an embodiment of the present invention;

[0038] Figure 4 Schematic diagram of the system structure of the adaptive full-view PCB image acquisition and QR code recognition and archiving system provided by an embodiment of the present invention;

[0039] Figure 5 This is a software architecture flow chart of the adaptive full-view PCB image acquisition and QR code recognition and archiving system provided by an embodiment of the present invention;

[0040] Figure 61 is a flow chart of an adaptive full-view PCB image acquisition and QR code recognition and archiving method provided by an embodiment of the present invention;

[0041] Figure 7 This is a software interface diagram of the adaptive full-view PCB image acquisition and QR code recognition and archiving system provided by an embodiment of the present invention.

[0042] Reference numerals:

[0043] 1. Ball screw guide rail; 2. Stepper motor; 3. Guide rail slider; 4. Screw slider; 5. Slider connecting plate; 6. Right-angle adapter bracket; 7. Platform plate; 8. Industrial camera; 9. Camera lens; 10. Camera gripper; 11. Bracket pole; 12. Light source bracket; 13. Strip light source. DETAILED DESCRIPTION

[0044] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0045] Example

[0046] like Figures 1 to 7 , adaptive full-view PCB image acquisition and QR code recognition archiving system, including hardware and software ends;

[0047] The hardware side includes a visual system module, a light source controller, a motion lifting module, a motion controller and an edge computer; the software side includes an image acquisition module, a detection and recognition module, a motion control module and a data decision module;

[0048] The motion lifting module is provided with a lifting structure; the visual system module is provided with an industrial camera 8 and a bar light source 13, and PCB image acquisition and QR code image acquisition are performed through the motion lifting module and the visual system module; the light source controller is used to control the bar light source 13; the motion controller is used to drive the stepping motor 2 to rotate and convert the motion mode through the edge computer;

[0049] The motion control module controls the motion of the lifting structure in the motion lifting module through a motion controller; the image acquisition module performs image acquisition and parameter configuration by controlling the visual system module; the detection and recognition module deploys a two-layer deep learning model in the software architecture of the edge computer, detects each QR code and stores information, and optionally creates multi-level folders and updates the full-view image save path.

[0050] The data decision module is used to store images from various perspectives; the edge computer makes decisions and manages data by controlling each module.

[0051] Specifically, the motion lifting module is used to adapt to the QR code recognition and image acquisition of PCBs of different three-dimensional sizes, including coping with the problems of different lengths, widths, heights, fields of view and depths of field at different viewing angles.

[0052] The lifting structure of the motion lifting module includes a ball screw guide rail 1, a stepping motor 2, a guide rail slider 3, a screw slider 4, a slider connecting plate 5, a right-angle adapter 6 and a platform plate 7;

[0053] The stepper motor 2 is arranged on the upper part of the ball screw guide rail 1 to drive the ball screw guide rail 1. The guide rail slider 3 is arranged on the ball screw guide rail 1. The screw slider 4 is installed on the ball screw guide rail 1 through the guide rail slider 3 and moves up and down along the ball screw guide rail 1. The screw slider 4 is connected to two right-angle adapters 6 through a slider connecting plate 5. The right-angle adapter 6 is connected to the platform plate 7. The platform plate 7 reciprocates up and down with the screw slider 4.

[0054] The two ball screw guide rails 1 limit the unnecessary degrees of freedom of the screw slider 4 except for the linear reciprocating motion. The screw slider 4 has a built-in screw nut and is installed on the ball screw guide rail 1 to realize the linear up and down reciprocating motion of the entire platform plate 7 in a stroke of 500mm, ensuring that the imaging target position meets the focal length and depth of field requirements of the camera lens 9, and performing high-quality imaging.

[0055] The vision system module is used to obtain the QR code image on the PCB for recognition and to collect high-resolution images of the PCB target from multiple perspectives for archiving and storage.

[0056] The visual system module includes an industrial camera 8, a camera lens 9, a camera gripper 10, a support pole 11, a light source support 12 and a strip light source 13;

[0057] The camera lens 9 is located on the industrial camera 8 and faces the platform plate 7; the industrial camera 8 is set on the camera clamp 10, the camera clamp 10 is connected to the bracket pole 11, and the camera clamp 10 can be adjusted; the bracket pole 11 is also provided with a light source bracket 12, the light source bracket 12 is located below the camera clamp 10, and the light source bracket 12 can be adjusted, and two strip light sources 13 are provided on the light source bracket 12, which uniformly illuminate the platform plate 7 from both sides.

[0058] Two bar-shaped light sources 13 are used to evenly illuminate the PCB object on platform plate 7 from either side, ensuring that the industrial camera 8 can clearly capture the PCB's feature points and the QR code object. This design ensures that even the smallest QR code object can be captured with sufficient pixels for high-quality images, providing stability and efficiency for subsequent inspection and recognition, while also ensuring high-quality archival images from all angles.

[0059] The ball screw guide rail 1 is arranged parallel to the bracket upright 11, and the visual system module and the motion lifting module are located on the same board. A light source controller, a motion controller and an edge computer are provided below the visual system module and the motion lifting module.

[0060] The light source controller drives and controls two bar light sources 13, adjusts the light source power through digital signals to control the light source brightness and adapt to different working scenarios; the motion controller drives and controls the rotation of the stepper motor 2, receives the pulse level signal from the edge computer, drives the rotation of the stepper motor 2, and converts the rotational motion into the linear reciprocating motion of the screw slider 4 through the ball screw guide 1.

[0061] Edge computers are used to control and coordinate the work of various modules of the system, run system software, make decisions based on data, and perform data management and visualization, serving as the brain of the system.

[0062] The image acquisition module is used to control the shooting process of the industrial camera 8 from the visual system module and read the real-time image stream of the industrial camera 8; specifically, it includes:

[0063] Initialize the industrial camera 8, configure the parameters of the industrial camera 8 for QR code detection, trigger image acquisition according to the instructions sent by the edge computer, digitize the image data and transmit it to the edge computer for subsequent detection and recognition operations;

[0064] After the detection and recognition is completed, the parameters of the industrial camera 8 are switched to the configuration corresponding to the full-view PCB image acquisition, and the image stream is transmitted to the edge computer for subsequent batch archiving and storage.

[0065] The detection and identification module detects the QR code object based on the collected PCB image and identifies the information in the QR code object;

[0066] Two layers of deep learning models are deployed in the software architecture of the edge computer. The first layer of the model first detects and expands the segmentation of small QR code objects in the large field of view to achieve the dynamic ROI effect under the premise of ensuring the completion of the QR code object; the second layer of the model recognizes the content of the separated QR code objects and obtains the string content information of the QR code objects; this QR code recognition and detection operation is applicable to the situation where there are multiple QR code objects in the field of view.

[0067] The detection and identification module identifies and detects one or more two-dimensional code objects, and stores the detection and character string content information of each two-dimensional code object.

[0068] Detect and store the content information of each QR code object, and then choose to create a new multi-level folder for archiving.

[0069] The string content information of the QR code object is checked for legitimacy against the preset rules. Under the premise of legitimacy, the string content information is processed through regularized matching to obtain the model and number of the product corresponding to the PCB image. Based on the string content information, multi-level folders are created and the full-view image save path is updated.

[0070] The motion control module receives the level pulse information converted by the edge computer based on the lifting distance information, and controls the lifting structure of the motion lifting module to perform linear reciprocating motion through the motion controller and the stepper motor 2;

[0071] The data decision module is used to store the images and names of each perspective corresponding to the PCB image, store them in the image list and name list, perform legality detection on the image saving path, and automatically save the full-perspective images locally in batches through software triggering if they are legal. Then, each process is recorded and the results are visualized on the software interface.

[0072] The adaptive full-view PCB image acquisition and QR code recognition and archiving method applies the adaptive full-view PCB image acquisition and QR code recognition and archiving system described in any one of the above. The specific method is as follows:

[0073] The user places the PCB to be processed on the platform plate 7, which is then raised or lowered to the optimal height for QR code detection. The industrial camera 8 is configured for detection and recognition based on the current parameter information. The detection and recognition module detects and recognizes the QR code. This operation is repeated for all QR code objects in the field of view. All QR code objects are detected and recognized, and the results of multiple QR code detections are annotated and visualized.

[0074] The user selects a QR code object for creating a new multi-level folder for archiving. A multi-level folder is created based on the model and number information of the object, and the save path for the updated full-view image is generated. The current parameter information of the industrial camera 8 is switched to the full-view acquisition configuration. The platform adaptively rises and falls to the appropriate height according to the three-dimensional size of the PCB. The software-side operation collects and names images from different perspectives. Finally, the software triggers the automatic saving of batch images, completing the task of automatically archiving PCB objects according to model and number.

[0075] The present invention designs an adaptive full-view PCB image acquisition and QR code recognition archiving system and method to solve the problem of image acquisition and QR code recognition relying on traditional methods in the existing PCB manufacturing process, thereby improving production efficiency and accuracy.

[0076] We designed and built a hardware automation system, combined with a lifting platform, that can adaptively recognize and inspect QR codes on PCBs of varying three-dimensional dimensions, as well as capture images from various angles. This full-view image acquisition method ensures that every detail of the PCB is accurately captured, avoiding the limited field of view and depth of field associated with traditional fixed platforms and cameras, significantly improving the comprehensiveness and accuracy of inspections.

[0077] A high-precision recognition algorithm for small QR codes has been proposed, capable of identifying the content of a 5x5mm QR code within a 300x300mm field of view. The algorithm comprises a two-layer deep learning model. The first layer detects and segments the small QR code within the larger field of view, ensuring the integrity of the QR code object and achieving a dynamic ROI effect. The second layer then identifies the content of the separated QR code object, obtaining the string content information. This algorithm differs from existing QR code recognition algorithms in that it has a wider detection range, high recognition accuracy, and stability.

[0078] The present invention targets the full-angle image acquisition and QR code content recognition and archiving tasks of PCBs. Traditional manual processing methods take an average of 20 seconds to complete the operations of scanning codes, creating multi-level folders for classification, and storing acquired images. Using this system, full-angle image acquisition of a PCB can be completed within 5 seconds and automatically archived according to the product model and number of the QR code content. This not only avoids the instability and missed detection caused by manual intervention, but also improves detection efficiency.

[0079] This system is highly adaptable and portable: on the hardware side, the platform plate 7, camera gripper 10, and light source bracket 12 can be re-used by adjusting their structures according to the three-dimensional dimensions of the specific PCB being inspected. This system is also applicable to all image acquisition and archiving needs based on QR code recognition other than PCBs. On the software side, the software architecture does not need to be adjusted; by modifying the regularized matching rules according to different QR code rules, the multi-level folder generation logic can be changed. The QR code detection model and recognition model retain the training interface. For new objects whose accuracy does not meet the requirements, it is only necessary to train based on the existing images, switch the loaded model, and then migrate the recognition object, realizing software-side reuse.

[0080] The software architecture integrates a data communication interface and archives full-view PCB images, facilitating subsequent statistical analysis and processing tasks such as traceability and batch upload management to the ERP system.

[0081] Based on the disclosure and teachings of the above description, those skilled in the art may also make changes and modifications to the above embodiments. Therefore, the present invention is not limited to the specific embodiments disclosed and described above, and modifications and variations of the present invention should also fall within the scope of protection of the claims of the present invention. In addition, although certain specific terms are used in this description, these terms are only for convenience of description and do not constitute any limitation to the present invention.

Claims

1. Adaptive full-view PCB image acquisition and QR code recognition archiving system, including hardware and software; characterized by: The hardware side includes a visual system module, a light source controller, a motion lifting module, a motion controller and an edge computer; the software side includes an image acquisition module, a detection and recognition module, a motion control module and a data decision module; The motion lifting module is provided with a lifting structure; the visual system module is provided with an industrial camera and a bar light source, and PCB images and QR code images are acquired through the motion lifting module and the visual system module; the light source controller is used to control the bar light source; the motion controller is used to drive the stepper motor to rotate and convert the motion mode through the edge computer; The motion control module controls the motion of the lifting structure in the motion lifting module through a motion controller; the image acquisition module performs image acquisition and parameter configuration by controlling the visual system module; the detection and recognition module deploys a two-layer deep learning model in the software architecture of the edge computer, detects each QR code and stores information, and optionally creates multi-level folders and updates the full-view image save path. The data decision module is used to store images from various perspectives; the edge computer makes decisions and manages data by controlling each module.

2. The adaptive full-view PCB image acquisition and QR code recognition and archiving system according to claim 1 is characterized in that: The lifting structure of the motion lifting module includes a ball screw guide rail, a stepping motor, a guide rail slider, a screw slider, a slider connecting plate, a right-angle adapter and a platform plate; The stepper motor is arranged on the upper part of the ball screw guide rail to drive the ball screw guide rail. The guide rail slider is arranged on the ball screw guide rail. The screw slider is installed on the ball screw guide rail through the guide rail slider and moves up and down along the ball screw guide rail. The screw slider is connected to two right-angle adapters through a slider connecting plate. The right-angle adapter is connected to the platform plate. The platform plate reciprocates up and down with the screw slider.

3. The adaptive full-view PCB image acquisition and QR code recognition and archiving system according to claim 2 is characterized in that: The visual system module includes an industrial camera, a camera lens, a camera gripper, a bracket pole, a light source bracket and a bar light source; The camera lens is located on the industrial camera and faces the platform board; the industrial camera is set on the camera clamp, the camera clamp is connected to the bracket pole, and the camera clamp is adjustable; a light source bracket is also provided on the bracket pole, the light source bracket is located below the camera clamp, and the light source bracket is adjustable, and two strip light sources are provided on the light source bracket, which uniformly illuminate the platform board from both sides.

4. The adaptive full-view PCB image acquisition and QR code recognition and archiving system according to claim 3 is characterized in that: The ball screw guide rail is arranged parallel to the bracket upright, the visual system module and the motion lifting module are located on the same board, and a light source controller, a motion controller and an edge computer are provided below the visual system module and the motion lifting module.

5. The adaptive full-view PCB image acquisition and QR code recognition and archiving system according to claim 4 is characterized in that: The light source controller drives and controls two bar-shaped light sources, and adjusts the light source power through digital signals to control the brightness of the light source; the motion controller drives and controls the rotation of the stepper motor, receives pulse level signals from the edge computer, drives the rotation of the stepper motor, and converts the rotational motion into linear reciprocating motion of the screw slider through the ball screw guide.

6. The adaptive full-view PCB image acquisition and QR code recognition and archiving system according to claim 1 is characterized in that: The image acquisition module is used to control the shooting process of the industrial camera from the visual system module and read the real-time image stream of the industrial camera; specifically, it includes: Initialize the industrial camera, configure the parameters of the industrial camera for QR code detection, trigger image acquisition according to the instructions sent by the edge computer, digitize the image data and transmit it to the edge computer for subsequent detection and recognition operations; After detection and recognition are completed, the industrial camera parameters are switched to the configuration corresponding to full-view PCB image acquisition, and the image stream is transmitted to the edge computer for subsequent batch archiving and storage.

7. The adaptive full-view PCB image acquisition and QR code recognition and archiving system according to claim 1 is characterized in that: The detection and identification module detects the QR code object based on the collected PCB image and identifies the information in the QR code object; The edge computer's software architecture deploys two layers of deep learning models. The first layer detects and segments small QR code objects within a large field of view. The second layer identifies the content of the separated QR code objects and obtains the string content information of the QR code objects. The detection and identification module identifies and detects one or more two-dimensional code objects, and stores the detection and character string content information of each two-dimensional code object.

8. The adaptive full-view PCB image acquisition and QR code recognition and archiving system according to claim 7 is characterized in that: The string content information of the QR code object is checked for legitimacy against preset rules, and the string content information is processed through regularized matching to obtain the model and number of the product corresponding to the PCB image. Based on the string content information, multi-level folders are created and the full-view image save path is updated.

9. The adaptive full-view PCB image acquisition and QR code recognition and archiving system according to claim 8, characterized in that: The motion control module receives the level pulse information converted by the edge computer based on the lifting distance information, and controls the lifting structure of the motion lifting module to perform linear reciprocating motion through the motion controller and the stepping motor; The data decision module is used to store the images and names of each perspective corresponding to the PCB image, and after checking the legitimacy of the image saving path, batch save the full-perspective images locally, and then record each process and visualize the results.

10. Adaptive full-view PCB image acquisition and QR code recognition archiving method, characterized in that: The adaptive full-view PCB image acquisition and QR code recognition and archiving system according to any one of claims 1 to 9 is applied, and the specific method is as follows: The user places the PCB to be processed on the platform board, which is then raised or lowered to the height required for QR code detection. The industrial camera is configured for detection and recognition based on the current parameter information. The detection and recognition module detects and recognizes the QR code, detects and recognizes all QR code objects, and annotates and visualizes the results of multiple QR code detections. The user selects a QR code object for creating a new multi-level folder for archiving. Based on the model and number information of the selected QR code object, a multi-level folder is created, a save path for the updated full-view image is generated, and the current parameter information of the industrial camera is switched to the full-view acquisition configuration. The platform plate is raised or lowered to the corresponding height according to the size of the PCB. The software end collects and names images from different perspectives, and then triggers batch saving of images to complete the archiving task of the PCB based on model and number.

Images