Chip character detection system based on neural network

Abstract

The invention is suitable for the technical field of industrial machine vision and automatic detection, and provides a neural network-based chip character detection system, which comprises a continuous conveying device, a system controller and a collaborative imaging subsystem, the collaborative imaging subsystem comprises a plurality of visual imaging units; the system controller is configured to operate an imaging scheduling and optimization neural network model; the cooperative imaging subsystem is configured to execute the following adaptive cooperative detection process, and the system achieves the purpose of distributing appropriate visual imaging units in real time according to the conveying speed and the load through imaging scheduling and optimization of a neural network model, supports two strategies of time-sharing trigger series detection and follow-up synchronous detection, and improves the detection accuracy. Different production rhythms and detection requirements are adapted, and the flexibility and efficiency of the system are improved; the system integrates a fixed visual imaging unit and a movable visual imaging unit, the optimal imaging mode can be selected for different detection tasks, all-directional and high-precision detection is achieved, and limitation of a single sensor is avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of industrial machine vision and automatic detection technology, more particularly, it relates to a chip character detection system based on neural network. BACKGROUND

[0002] In the semiconductor manufacturing and packaging testing industry, the characters (such as model, batch number, trademark, etc.) on the surface of the chip are important basis for product traceability, quality control and authenticity identification. Traditional chip character detection mainly relies on manual inspection or fixed optical character recognition system, which has low detection efficiency, poor consistency, is easily disturbed by environmental light, and has difficulty in recognizing complex background and low contrast characters. With the development of chip packaging towards miniaturization and high integration, the character size is further reduced and the printing process is diversified, so the traditional detection method has been difficult to meet the production demand of high precision and high speed.

[0003] There are currently automatic detection schemes on the market that use a single vision sensor combined with a fixed light source, but they usually lack the ability to adapt to different imaging conditions (such as glare, shadow, motion blur, etc.). Especially in a continuous conveyor production line, the chip position, posture and lighting conditions may change at any time, resulting in unstable imaging quality and affecting the accuracy of character recognition. In addition, the fixed shooting mode is difficult to balance detection speed and image quality, especially on high-speed assembly lines, which is prone to missed detection or false detection.

[0004] Therefore, an intelligent detection system that can adapt to complex production line environments, implement dynamic imaging scheduling and adaptive optimization is urgently needed to improve the robustness, efficiency and accuracy of chip character detection. SUMMARY

[0005] In view of the deficiencies of the prior art, the purpose of the present application is to provide a chip character detection system based on neural network, which realizes imaging scheduling and optimization of neural network model, and according to the conveying speed and load, it realizes real-time allocation of appropriate vision imaging units, supports two strategies of time-sharing trigger series detection and follow-up synchronous detection, adapts to different production rhythm and detection demand, and improves the flexibility and efficiency of the system.

[0006] To achieve the above purpose, the present application provides the following technical scheme: A chip character detection system based on neural network, comprising a continuous conveying device, a system controller and a cooperative imaging subsystem; The cooperative imaging subsystem comprises a plurality of vision imaging units; The system controller is configured to run an imaging scheduling and optimization neural network model; The cooperative imaging subsystem is configured to perform the following adaptive cooperative detection process: An imaging scheduling stage: the system controller dynamically allocates at least one dedicated visual imaging unit for the chip trays flowing through different physical sections and generates an initial imaging scheduling strategy through an imaging scheduling and optimization neural network model according to the speed and load of the continuous conveying device; The initial imaging scheduling strategy is one of the following detection strategies or a combination thereof: Time-sharing trigger serial detection strategy: the visual imaging unit is a fixed visual imaging unit; the fixed visual imaging unit is controlled to sequentially take pictures of the same chip tray at different time points; Follow-up synchronous detection strategy: including a synchronization unit; the visual imaging unit is a mobile visual imaging unit; the synchronization unit is controlled to drive at least one mobile visual imaging unit to move synchronously with the corresponding chip tray and complete the shooting during the movement; Adaptive optimization stage: obtaining the chip images collected by the visual imaging unit in the imaging scheduling stage; inputting the chip images into the imaging scheduling and optimization neural network model, analyzing the image quality and identifying the imaging challenges of the chip characters by the model; generating optimization instructions based on the imaging challenges by the imaging scheduling and optimization neural network model; Cooperative decision-making stage: based on the final images obtained after the adaptive optimization stage, the system completes the final detection and judgment of the chip characters.

[0007] The application further provides that the optimization instructions include at least one of the following: Adjusting the illumination parameters of the current or subsequent visual imaging unit; Adjusting the camera parameters of the current or subsequent visual imaging unit; Correcting or regenerating the initial imaging scheduling strategy.

[0008] The application further provides that the continuous conveying device includes a conveying frame; four groups of rotating shafts are symmetrically and rotatably arranged between the inner walls of the conveying frame; an encoder is installed on the rotating shaft; a belt pulley is fixed symmetrically on the circumferential surface of the rotating shaft; a belt adapted to the belt pulley is drivingly arranged between the two belt pulleys; a plurality of accommodating grooves are uniformly formed on the outer wall of the belt; a accommodating cavity is formed by the two accommodating grooves coaxially; a servo motor is fixedly installed on the side surface of the conveying frame; the output end of the servo motor is fixedly connected with one end of the rotating shaft; first gears are fixedly arranged on the circumferential surfaces of the adjacent two rotating shafts and are in meshing engagement.

[0009] The application further provides that a supporting table is rotatably arranged in the conveying frame; a plurality of groups of chip trays are slidingly arranged on the surface of the supporting table; connecting columns adapted to the corresponding accommodating cavities are symmetrically fixed on the opposite side surfaces of the chip trays.

[0010] The application is further provided: the fixed visual imaging unit comprises a top plate fixedly installed on the top of the conveying frame; three groups of first industrial cameras are sequentially fixedly installed on the bottom surface of the top plate; a first lens is installed on the first industrial camera; a coaxial light source for character detection, a dome light source for appearance detection, and a low-angle light source for pin detection are respectively installed on the bottom of the three groups of first industrial cameras.

[0011] The application is further provided: the synchronous unit comprises a transmission shaft symmetrically and rotatably arranged on the conveying frame; conveying rollers are fixedly arranged on the circumferential surface of the transmission shaft; a conveying belt matched with the conveying rollers is rotatably arranged between the two conveying rollers; a second gear is fixedly arranged on one end of the transmission shaft; and a third gear matched with the second gear is fixedly arranged on the other end of the transmission shaft.

[0012] The application is further provided: the movable visual imaging unit comprises a mounting plate fixedly arranged on the surface of the conveying belt; an inner light shielding cylinder is fixedly arranged on the surface of the mounting plate; a second industrial camera is fixedly arranged on the surface of the mounting plate; a second lens is fixedly arranged on the second industrial camera; a pulsed light source is arranged in the inner light shielding cylinder; guide rods are symmetrically fixedly arranged on the surface of the mounting plate; and an outer light shielding cylinder is slidably arranged between the two guide rods.

[0013] The application is further provided: a sliding rod is fixedly arranged on the outer wall of the outer light shielding cylinder; an annular track slidably matched with the sliding rod is fixedly arranged on the conveying frame; inclined tracks connected with each other are downwardly and extendingly arranged at both ends of the bottom of the annular track; a straight track connected with the inclined tracks is arranged between the two inclined tracks; and pressure switches for starting and stopping the second industrial camera are fixedly arranged on the surface of the straight track close to both ends thereof.

[0014] The application has the following advantages:

[0015] 1. The application adjusts and optimizes the imaging scheduling and the neural network model, and appropriately divides the visual imaging unit according to the conveying speed and the load in real time, supports time-sharing triggering series detection and follow-up synchronous detection, adapts to different production rhythms and detection requirements, and improves the flexibility and efficiency of the system.

[0016] 2. The application analyzes imaging challenges (such as uneven illumination, motion blur, and reflection interference) based on real-time image acquisition by a neural network, and dynamically adjusts the light source, camera parameters, or scheduling strategy to improve image quality from the source and enhance the robustness of character recognition.

[0017] 3. The application integrates fixed and movable visual imaging units, and can select the best imaging mode for different detection tasks (characters, appearance, and pins) to realize omnidirectional and high-precision detection and avoid the limitations of a single sensor.

[0018] 4. The application realizes continuous conveying and synchronous shooting of the chip tray by mechanical and control cooperation of the continuous conveying device and the imaging unit, supports non-stop detection, and greatly improves the production line throughput. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the structure of a chip character detection system based on a neural network according to the present invention.

[0020] Figure 2 This is a flowchart illustrating the adaptive collaborative detection process of the present invention.

[0021] Figure 3 This is a schematic diagram of the continuous conveying device of the present invention.

[0022] Figure 4 For the present invention Figure 3 A structural diagram from a frontal viewpoint.

[0023] Figure 5 This is a schematic diagram of the structure of the fixed visual imaging unit of the present invention.

[0024] Figure 6 This is a schematic diagram of the detection structure in Embodiment 2 of the present invention.

[0025] Figure 7 For the present invention Figure 6 A perspective that faces the issue directly.

[0026] Figure 8 For the present invention Figure 7 Enlarged view of region A.

[0027] Figure 9 This is a schematic diagram of the detection structure in Embodiment 3 of the present invention.

[0028] Figure 10 For the present invention Figure 9 Enlarged view of region B.

[0029] In the diagram: 1. Continuous conveying device; 2. Chip tray; 3. Synchronization unit; 4. Conveyor frame; 5. Rotary shaft; 6. Pulley; 7. Belt; 8. Receiving groove; 9. Receiving cavity; 10. Servo motor; 11. Sliding rod; 12. Circular track; 13. Support platform; 14. Connecting column; 15. Top plate; 16. First industrial camera; 17. First lens; 18. Coaxial light source; 19. Dome light source; 20. Low-angle light source; 21. Drive shaft; 22. Conveyor roller; 23. Conveyor belt; 24. Inclined track; 25. Straight track; 26. Pressure switch; 27. Mounting plate; 28. Inner light shield; 29. ​​Second industrial camera; 30. Second lens; 31. Pulse light source; 32. Guide rod; 33. Outer light shield. Detailed Implementation

[0030] It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other in the case of no conflict. The present application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0031] It should be noted that, unless otherwise specified, all technical and scientific terms used in the present application have the same meaning as commonly understood by those skilled in the art to which the present application belongs.

[0032] In the present application, unless otherwise stated, the orientation such as "up, down" is generally directed to the direction shown in the drawings, or to the vertical, perpendicular or gravity direction; similarly, for the convenience of understanding and description, "left, right" is generally directed to the left and right shown in the drawings; "inner, outer" refers to the inner and outer relative to the contour of each component itself, but the above orientation words are not used to limit the present application.

[0033] Embodiment one, please refer to Figures 1-10 The present application provides the following technical solutions: A chip character detection system based on a neural network, specifically comprising a continuous conveying device 1, a system controller, and a cooperative imaging subsystem; the cooperative imaging subsystem comprises a plurality of visual imaging units; The system controller is configured to run an imaging scheduling and optimization neural network model; The cooperative imaging subsystem is configured to perform the following adaptive cooperative detection process: Imaging scheduling stage: the system controller dynamically allocates at least one dedicated visual imaging unit to the chip tray 2 flowing through different physical sections according to the speed and load of the continuous conveying device 1 through the imaging scheduling and optimization neural network model, and generates an initial imaging scheduling strategy; The initial imaging scheduling strategy is one of the following detection strategies or a combination thereof: Time-sharing trigger series detection strategy: the visual imaging unit is a fixed visual imaging unit; the fixed visual imaging unit is controlled to sequentially shoot the same chip tray 2 at different time points; Follow-up synchronous detection strategy: including a synchronization unit 3; the visual imaging unit is a mobile visual imaging unit; the synchronization unit 3 is controlled to drive at least one mobile visual imaging unit and the corresponding chip tray 2 to move synchronously, and the shooting is completed during the movement; Adaptive optimization stage: the chip images collected by the visual imaging unit in the imaging scheduling stage are obtained; the chip images are input into the imaging scheduling and optimization neural network model, the model analyzes the image quality and identifies the imaging challenges of the chip characters; based on the imaging challenges, the imaging scheduling and optimization neural network model dynamically generates optimization instructions; Cooperative decision-making stage: based on the final image obtained after adaptive optimization, the system completes the final detection and judgment of the chip characters.

[0034] The optimization instruction comprises at least one of: adjusting the lighting parameters of the current or subsequent visual imaging unit; adjusting the camera parameters of the current or subsequent visual imaging unit; correcting or regenerating the initial imaging scheduling strategy.

[0035] The continuous conveying device 1 comprises a conveying frame 4; four groups of rotating shafts 5 are symmetrically and rotatably arranged between the inner walls of the conveying frame 4; a belt pulley 6 is fixedly arranged on the symmetric side of the rotating shaft 5; a matched belt 7 is arranged in transmission between the two belt pulleys 6; a plurality of accommodating grooves 8 are uniformly arranged on the outer wall of the belt 7; the two accommodating grooves 8 on the same axis form an accommodating cavity 9; a servo motor 10 is fixedly installed on the side surface of the conveying frame 4; the output end of the servo motor 10 is fixedly connected with the end of one rotating shaft 5; the symmetrically meshed first gears are fixedly arranged on the circumferential side of the adjacent two rotating shafts 5.

[0036] A supporting table 13 is arranged in the conveying frame 4; a plurality of chip trays 2 are slidingly arranged on the surface of the supporting table 13; the connecting columns 14 symmetrically fixed on the opposite sides of the chip tray 2 are matched with the corresponding accommodating cavities 9.

[0037] The working principle of the embodiment is as follows: By placing a plurality of chip trays 2 on the supporting table 13 in sequence, and pushing the outermost chip tray 2, the chip trays 2 are moved towards the belt 7; the corresponding rotating shaft 5 is driven to rotate by the control of the servo motor 10, and the two meshed first gears are driven to rotate, thereby driving the lower rotating shaft 5 to synchronously and reversely rotate, and further driving the two groups of upper and lower belt 7 to synchronously and reversely rotate; the connecting columns 14 at the two ends of the chip tray 2 closest to the belt 7 are clamped into the corresponding accommodating cavities 9, thereby completing the continuous conveying of the plurality of chip trays 2 in sequence.

[0038] The imaging scheduling and optimization neural network model is a multi-task deep learning model, which is deployed on an embedded system or an industrial control computer, and has the dual capabilities of dynamic scheduling decision and image quality optimization analysis. The model adopts an encoder-multi-branch decoder structure, which specifically comprises: Feature extraction encoder: a convolutional neural network (such as ResNet, EfficientNet, etc.) is used to extract multi-level features from the input chip image, and output feature maps containing global and local information.

[0039] Dispatch decision branch: based on the features from the encoder output, combined with real-time state information from the conveyor (such as speed, position, load) through a fully connected layer with attention mechanism, output the type of vision imaging unit (fixed / movable) and its scheduling strategy (time-triggered / synchronous following) that the current chip tray 2 should be assigned.

[0040] Imaging analysis branch: image quality assessment and imaging challenge identification on the same feature map, output includes but not limited to: Image blur score; light uniformity analysis; glare area detection; character region contrast evaluation; motion artifact identification.

[0041] Model training and optimization The neural network is trained through a combination of end-to-end supervised learning and reinforcement learning: Supervised learning phase: using a labeled chip image dataset, the model is trained to identify image quality problems and imaging challenges, and to learn to generate optimization instructions (such as light source parameter adjustment, camera parameter suggestion) under different imaging conditions.

[0042] Reinforcement learning phase: in a simulated environment, the model learns how to make optimal scheduling decisions in a dynamic production line environment through interaction with virtual conveyors and imaging units to maximize detection success rate and throughput.

[0043] Collaboration mechanism with character recognition neural network The system can integrate a dedicated character recognition neural network (such as CRNN, Transformer-based OCR model) for character detection and recognition on the optimized final image. This model can be deployed in series or parallel with the imaging scheduling model, supporting the following two working modes: Series mode: after the imaging scheduling model outputs the optimized image, it is directly input to the character recognition network for final judgment.

[0044] Parallel collaboration mode: the two models share the feature extraction layer, realizing the synchronization of imaging optimization and character recognition, further improving the real-time performance of the system.

[0045] Real-time inference and deployment A lightweight inference engine (such as TensorRT, OpenVINO, etc.) is deployed in the system controller to support real-time running of the neural network on embedded devices or industrial workstations. The model supports online learning and incremental updating, and can continuously optimize the scheduling and imaging strategy according to the actual data of the production line.

[0046] By imaging scheduling and optimizing the neural network model, appropriate visual imaging units are allocated in real time according to the conveying speed and load, supporting time-sharing triggering serial detection and follow-up synchronous detection strategies, adapting to different production rhythms and detection needs, and improving system flexibility and efficiency; based on real-time collected images, neural network analyzes imaging challenges (such as uneven lighting, motion blur, and light interference), and dynamically adjusts light source, camera parameters or scheduling strategy to improve image quality from the source and enhance character recognition robustness; the system integrates fixed and mobile visual imaging units, which can select the best imaging mode for different detection tasks (characters, appearance, and pins) to achieve all-around and high-precision detection, avoiding the limitations of a single sensor.

[0047] Embodiment two, please refer to Figures 1-5 This embodiment two is improved on the basis of embodiment one, specifically, the shaft 5 is provided with an encoder; the fixed visual imaging unit includes a top plate 15 fixedly installed on the top of the conveying frame 4; the bottom surface of the top plate 15 is sequentially fixedly provided with three groups of first industrial cameras 16; the first industrial cameras 16 are provided with first lenses 17; the bottom of the three groups of first industrial cameras 16 is respectively provided with a coaxial light source 18 for character detection, a dome light source 19 for appearance detection, and a low-angle light source 20 for pin detection.

[0048] Working principle of this embodiment two: The coaxial light source 18 is designed for detecting small concave-convex or engraved characters (such as chip characters) on smooth surfaces, and its core is that the light path is parallel to the camera optical axis.

[0049] Core components: a beam splitter, a semi-transparent mirror placed at a 45-degree angle in front of the camera lens; the LED light source is horizontally injected from the side; the light is reflected vertically downward after passing through the beam splitter, and irradiates to the product surface, and the light reflected vertically from the surface passes through the beam splitter again and directly enters the camera lens.

[0050] It is an independent assembly, usually fixed directly to the front end of the camera lens through a ring or mounting bracket, and the entire "camera-lens-coaxial light source" is installed as a whole module on the bracket.

[0051] Chip detection effect: the smooth chip surface reflects the light vertically back to the camera, showing bright white; rough concave or convex characters undergo diffuse reflection, only a small amount of light returns to the camera, showing dark black.

[0052] Effect: it forms a black and white image with extremely high contrast, which is the preferred light source for character reading.

[0053] The dome light source 19 is designed to eliminate reflections and shadows, providing shadow-free uniform illumination.

[0054] Installation structure: A half-sphere cover with inner wall of diffuse reflection material.

[0055] LED light source: usually installed on the side or top of the dome, light rays are directed to the inner wall of the hemisphere.

[0056] Working principle: light rays are evenly and softly irradiated to the detected object from all angles after multiple diffuse reflections on the inner wall of the hemisphere.

[0057] Dome light source 19 itself is a separate, larger fixed device, and the industrial camera shoots downward from the inside of the dome through the opening at the top of the dome. The dome light is fixed on the rack, completely covering the chip tray 2 to be detected inside it.

[0058] Chip detection function: Eliminate reflection: greatly weaken the specular reflection of the smooth surface of the chip and the pins.

[0059] Eliminate shadows: achieve shadowless lighting, so that every detail on the surface of the chip is evenly illuminated.

[0060] Effect: obtain a whole and clear image, which is very suitable for overall appearance detection and color recognition.

[0061] Low-angle light source 20 is used to highlight the three-dimensional profile and height difference of the object.

[0062] Installation structure: Long strip LED light bead array.

[0063] When installed, the light rays graze the surface of the measured object at a very small angle (almost parallel).

[0064] Installation method: Directly installed around the camera lens: used as a low-angle ring light, the light rays are emitted at a low angle from around the lens.

[0065] Independent installation: one or more strip lights are fixed separately on the rack on both sides of the camera, and the irradiation angle is adjusted accurately to make it almost parallel to the chip surface.

[0066] Dark field illumination: an extreme form of low-angle light, with a lower installation angle, so that the light reflected by the flat surface cannot enter the lens at all (black), and only the uneven edges can scatter the light into the lens (white).

[0067] Chip detection function: Highlight the pins: the pins of the chip will block the light, forming a bright edge on the other side, while the bottom of the pins or the position of the missing pins will form a shadow.

[0068] Scratch / unevenness detection: any slight surface fluctuation will produce obvious light and dark differences due to light irradiation.

[0069] Effect: convert height information into gray scale information, mainly used for pin detection, warping detection, scratch detection, etc.

[0070] By installing a high-quality encoder on the rotating shaft 5, a unified position clock signal is generated. When the first industrial camera 16 takes a picture at time T, the system controller needs to accurately calculate which part of which chip tray 2 the picture belongs to according to the encoder position and the conveyor belt speed, and correctly associate the image data to the detection task of the chip tray 2, so as to realize "non-stop detection" of the chips inside the chip tray 2 and improve the detection efficiency and effect.

[0071] Embodiment three, please refer to Figure 1 、 Figure 2 、 Figures 6-10 , this embodiment three is improved on the basis of embodiment one, specifically, the synchronization unit 3 includes a transmission shaft 21 symmetrically penetrating and arranged on the conveying frame 4; the transmission shaft 21 is fixed with conveying rollers 22 on the side surface; the conveying belt 23 matched with the conveying rollers 22 is arranged in transmission between the two conveying rollers 22; a second gear is fixed at the end of a group of transmission shafts 21; a third gear matched with the second gear is fixed at the end of a group of rotating shafts 5.

[0072] The mobile visual imaging unit includes a mounting plate 27 fixed on the surface of the conveying belt 23; an inner light shielding cylinder 28 is fixed on the surface of the mounting plate 27; a second industrial camera 29 is installed on the surface of the mounting plate 27; a second lens 30 is installed on the second industrial camera 29; a pulse light source 31 is installed inside the inner light shielding cylinder 28; guide rods 32 are symmetrically fixed on the surface of the mounting plate 27; an outer light shielding cylinder 33 is slidingly arranged between the two groups of guide rods 32.

[0073] The outer light shielding cylinder 33 is fixed with a sliding rod 11; the conveying frame 4 is fixed with a ring track 12 slidingly matched with the sliding rod 11; the bottom of the ring track 12 extends downward to be provided with two inclined tracks 24 connected with each other; the two groups of inclined tracks 24 are connected with a straight track 25 connected with each other; the pressure switch 26 for controlling the start and stop of the second industrial camera 29 is fixed and installed on the surface of the straight track 25 near the two ends thereof.

[0074] The working principle of this embodiment three is as follows: By controlling the starting servo motor 10 to drive the corresponding upper and lower opposite belts 7 to rotate reversely in synchronization, the synchronous transmission of the conveying belt 23 is realized through the meshing and matching of the second gear and the third gear, and then each group of mobile visual imaging units and the chip tray 2 are synchronously transmitted, that is, each group of mobile visual imaging units can be synchronously transmitted with the corresponding chip tray 2 without waiting, thereby improving the detection efficiency.

[0075] When a set of mobile vision imaging units slides from the circular track 12 to the inclined track 24, the slide bar 11 is guided by the inclined track 24, causing it to descend during the movement. This drives the outer light-shielding tube 33 to slide down along the guide rod 32. When the corresponding mobile vision imaging unit slides from the inclined track 24 to the straight track 25, the outer light-shielding tube 33 covers the corresponding chip tray 2, effectively isolating ambient light and interference from adjacent workstations, ensuring a stable imaging environment, improving detection consistency, and enhancing detection quality. When the connecting post 14 on the mobile vision imaging unit presses the first set of pressure switches 26, the pressure switch 26 transmits a signal to the system controller. The system controller then controls the corresponding second industrial camera 29 and pulse light source 31 to start, enabling the outer light-shielding tube 33 to cover the corresponding chip tray 2 for detection. When the corresponding connecting post 14 presses the second set of pressure switches 26, the system controller controls the corresponding second industrial camera 29 and pulse light source 31 to shut down.

[0076] Obviously, the embodiments described above are merely some, not all, embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort should fall within the scope of protection of the present invention.

[0077] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0078] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in sequences other than those illustrated or described herein.

[0079] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0080] The above merely describes the preferred embodiments of the present application, and the protection scope of the present application is not limited to the above-described embodiments. Any technical solution falling within the concept of the present application shall fall within the protection scope of the present application. It should be noted that, for ordinary skilled persons in the art, some improvements and refinements without departing from the principles of the present application shall also be considered as falling within the protection scope of the present application.

Claims

1. A chip character detection system based on a neural network, characterized in that: Includes a continuous transport device (1), a system controller, and a collaborative imaging subsystem; The collaborative imaging subsystem includes multiple visual imaging units; The system controller is configured to run an imaging scheduling and optimization neural network model; The collaborative imaging subsystem is configured to perform the following adaptive collaborative detection process: Imaging scheduling stage: Based on the speed and load of the continuous conveying device (1), the system controller dynamically allocates at least one dedicated visual imaging unit to the chip trays (2) flowing through different physical sections through imaging scheduling and optimization neural network model, and generates an initial imaging scheduling strategy. The initial imaging scheduling strategy is one of the following detection strategies or a combination thereof: Time-division triggering serial detection strategy: The visual imaging unit is a fixed visual imaging unit; control the fixed visual imaging unit to take pictures of the same chip tray (2) at different time points in sequence; Follow-up synchronous detection strategy: including a synchronization unit (3); the visual imaging unit is a mobile visual imaging unit; the control synchronization unit (3) drives at least one mobile visual imaging unit to move synchronously with the corresponding chip tray (2) and completes the shooting during the movement; Adaptive optimization phase: Acquire chip images from the visual imaging unit during the imaging scheduling phase; input the chip images into the imaging scheduling and optimization neural network model, which analyzes the image quality and identifies the imaging challenge of chip characters; Based on imaging challenges, an imaging scheduling and optimization neural network model dynamically generates optimization instructions. Collaborative decision-making stage: Based on the final image obtained after the adaptive optimization stage, the system completes the final detection and judgment of the chip characters.

2. The chip character detection system based on a neural network according to claim 1, characterized in that: The optimization instructions include at least one of the following: Adjust the illumination parameters of the current or subsequent visual imaging units; Adjust the camera parameters of the current or subsequent visual imaging units; The initial imaging scheduling strategy is modified or regenerated.

3. The chip character detection system based on a neural network according to claim 2, characterized in that: The continuous conveying device (1) includes a conveying frame (4); four sets of rotating shafts (5) are symmetrically arranged through the inner wall of the conveying frame (4); an encoder is installed on the rotating shaft (5); pulleys (6) are symmetrically fixed on the circumferential side of the rotating shaft (5); a matching belt (7) is arranged between the two pulleys (6); several receiving grooves (8) are evenly opened on the outer wall of the belt (7); two coaxial receiving grooves (8) form a receiving cavity (9); a servo motor (10) is fixedly installed on the side of the conveying frame (4); the output end of the servo motor (10) is fixedly connected to the end of a rotating shaft (5); a first gear that meshes with each other is fixed on the circumferential side of two adjacent rotating shafts (5).

4. The chip character detection system based on a neural network according to claim 3, characterized in that: The conveyor (4) has a support platform (13) running through its interior; multiple chip trays (2) are slidably disposed on the surface of the support platform (13); the chip trays (2) have connecting columns (14) that are adapted to the corresponding receiving cavities (9) fixed symmetrically on opposite sides.

5. A chip character detection system based on a neural network according to claim 4, characterized in that: The fixed vision imaging unit includes a top plate (15) fixedly installed on the top of the conveyor (4); three sets of first industrial cameras (16) are fixedly installed on the bottom surface of the top plate (15); a first lens (17) is installed on the first industrial camera (16); and a coaxial light source (18) for character detection, a dome light source (19) for appearance detection, and a low-angle light source (20) for pin detection are respectively installed at the bottom of the three sets of first industrial cameras (16).

6. The chip character detection system based on a neural network according to claim 4, characterized in that: The synchronization unit (3) includes a drive shaft (21) symmetrically and rotatably mounted on the conveyor frame (4); a conveyor roller (22) is fixed on the circumferential side of the drive shaft (21); a conveyor belt (23) adapted to the two conveyor rollers (22) is driven between them; a second gear is fixed at the end of a set of drive shafts (21); a third gear that meshes with the second gear is fixed at the end of a set of rotating shafts (5).

7. A chip character detection system based on a neural network according to claim 6, characterized in that: The mobile vision imaging unit includes a mounting plate (27) fixed on the surface of the conveyor belt (23); an inner light shield (28) is fixed on the surface of the mounting plate (27); a second industrial camera (29) is mounted on the surface of the mounting plate (27); a second lens (30) is mounted on the second industrial camera (29); a pulse light source (31) is installed inside the inner light shield (28); guide rods (32) are symmetrically fixed on the surface of the mounting plate (27); and an outer light shield (33) is slidably arranged between the two sets of guide rods (32).

8. A chip character detection system based on a neural network according to claim 7, characterized in that: A sliding rod (11) is fixed on the outer wall of the outer light shield (33); a ring track (12) with sliding cooperation of the sliding rod (11) is fixedly installed on the conveyor frame (4); inclined tracks (24) are connected to each other at both ends of the bottom of the ring track (12); a straight track (25) is connected between the two sets of inclined tracks (24); pressure switches (26) for controlling the start and stop of the second industrial camera (29) are fixedly installed on the surface of the straight track (25) near both ends.

Images