Optical monitoring system and method for quality detection

Abstract

The application provides an optical monitoring system and method for quality detection, the system comprising: an imaging module comprising at least one camera and a 3D sensor for collecting image data and three-dimensional topography data of a dispensing area; an illumination module comprising at least two light sources with different incident angles or different polarization states for providing adjustable illumination to the dispensing area to enhance the contrast between the glue and the background; a synchronous control module configured to synchronously trigger the imaging module and coordinate the light source switching of the illumination module; and a processing module connected with the imaging module and the synchronous control module, configured to perform fusion processing on the synchronously collected image data and three-dimensional topography data, and judge the dispensing quality based on the fusion result. The technical scheme of the application can realize high-precision online automatic dispensing quality detection.

Description

Technical Field

[0001] This invention relates to the fields of machine vision and industrial automation inspection technology, and in particular to an optical monitoring system for quality inspection. Background Technology

[0002] Dispensing processes are widely used in industries such as electronic packaging, automotive manufacturing, and medical devices. The quality of dispensing (such as the width, height, continuity, and coverage of the adhesive path) directly affects the sealing performance, bond strength, and long-term reliability of the product. Traditional dispensing quality inspection mainly relies on manual visual inspection or contact measurement, which suffers from low efficiency, high subjectivity, and a high risk of missed detections.

[0003] With the development of machine vision technology, 2D camera-based vision inspection systems have been applied to quality inspection. However, these 2D vision systems have significant technical limitations. They can only acquire planar information of the adhesive path and cannot detect the height and total volume of the adhesive, which are crucial for sealing / bonding performance. This makes it difficult to effectively detect key defects such as "excessive adhesive" (too large volume) or "insufficient adhesive" (insufficient height). Secondly, when dealing with highly reflective, low-contrast, or curved workpieces, the imaging quality is unstable, and the edges of the adhesive path are difficult to extract accurately. Furthermore, for deformations caused by fluid characteristics such as collapse, backflow, and stringing after dispensing, a single static image cannot accurately represent the final curing state, especially making it difficult to distinguish between true "broken adhesive" and extremely fine continuous adhesive formed by "stringing."

[0004] To address the height measurement problem, various issues arise. For example, adhesive materials (such as transparent UV adhesives and highly reflective silicone) and complex workpiece surfaces can interfere with the measurement process, leading to large height data errors and unstable edge recognition. Furthermore, high-speed production lines demand extremely high inspection cycle times (typically requiring single-piece inspection time to be less than 0.5 seconds), and complex data processing methods struggle to meet real-time requirements.

[0005] Therefore, there is an urgent need for an online dispensing quality detection system and method that can overcome the above-mentioned defects, achieve high precision, high efficiency, high robustness, and adapt to complex industrial environments. Summary of the Invention

[0006] The purpose of this invention is to provide an optical monitoring system and method for quality inspection, which can achieve high-precision and robust online automated quality inspection by fusing image data and three-dimensional topography data, adjusting illumination according to working conditions, and precise synchronous control.

[0007] According to one aspect of the present invention, an optical monitoring system for quality inspection is provided, comprising: An imaging module, comprising at least one camera and a 3D sensor, for acquiring image data and three-dimensional topography data of the dispensing area; An illumination module comprising at least two light sources with different incident angles or different polarization states, for providing adjustable illumination to the dispensing area to enhance the contrast between the adhesive and the background; A synchronization control module is configured to synchronously trigger the imaging module and coordinate the light source switching of the illumination module; The processing module is connected to the imaging module and the synchronization control module, and is configured to perform fusion processing on the acquired image data and three-dimensional topography data, and judge the dispensing quality based on the fusion result.

[0008] According to some embodiments, the lighting module includes at least two of ring light, side light, and coaxial light; The synchronization control module is also configured to dynamically switch or combine different light sources based on the surface characteristics or colloid type of the product being tested.

[0009] According to some embodiments, the camera includes a 2D camera or a 3D camera; The 3D sensor includes a line laser profilometer, a structured light projector, or a binocular stereo vision camera.

[0010] According to some embodiments, an optical adjustment mechanism is also included for overlapping the field of view of the camera and the 3D sensor in the dispensing area.

[0011] According to some embodiments, the processing module performs fusion processing on the synchronously acquired image data and three-dimensional topography data, including: extracting two-dimensional contour information and three-dimensional height information of the adhesive path, and calculating the width, height and cross-sectional area of ​​the adhesive path based on the two-dimensional contour information and the three-dimensional height information.

[0012] According to some embodiments, the processing module is configured as follows: The imaging module continuously acquires multiple images of the moving product to obtain a series of consecutive image frames and / or a three-dimensional data sequence. Analyze the continuity of the adhesive path trajectory in the sequence; Based on the continuity analysis results, distinguish between glue breakage defects and / or visual pseudo-defects caused by glue flow.

[0013] According to some embodiments, the processing module is configured to perform localization and registration of the image data and three-dimensional topography data based on Mark points on the product; The processing module is further configured to divide the dispensing area into multiple sub-regions, and to preset independent qualification thresholds for the width and / or height of the adhesive path for each sub-region.

[0014] According to another aspect of the present invention, an optical monitoring method for quality inspection is provided, applied to the system described in any of the preceding claims, the method comprising: The imaging module acquires image information of the dispensing area under the selected lighting mode; The synchronous control module triggers the imaging module to synchronously acquire image data and three-dimensional morphological data of the dispensing area; The processing module fuses the synchronously acquired image data and three-dimensional topography data to obtain fused data containing two-dimensional contour information and three-dimensional height information of the adhesive path. Based on the fused data, the two-dimensional contour information and three-dimensional height information are compared with preset standards to determine whether the dispensing quality is qualified.

[0015] According to some embodiments, the step of triggering the imaging module through the synchronization control module to synchronously acquire image data and three-dimensional morphology data of the dispensing area includes: Multiple consecutive data acquisitions are performed during the product's movement to obtain a time-series data sequence; The inter-frame continuity of the adhesive path trajectory in the time-series data sequence is analyzed, and based on the continuity analysis results, it is determined whether there is an adhesive breakage defect in the adhesive path.

[0016] According to some embodiments, the method further includes: If a defect is identified, a corresponding dispensing process parameter compensation instruction is generated. The process parameter compensation command is sent to the dispensing equipment or production management system to adjust the dispensing process of subsequent products.

[0017] According to embodiments of this application, a camera and a 3D sensor are integrated into the same imaging module, overcoming the limitation of traditional pure 2D vision systems that can only measure the width / area of ​​a plane. By simultaneously acquiring high-resolution two-dimensional images and accurate three-dimensional topographic data, a comprehensive, high-precision, three-dimensional evaluation of dispensing quality is achieved.

[0018] By configuring an illumination module containing at least two light sources with different incident angles or polarization states, and combining it with a synchronous control module that can coordinate light source switching, the system possesses active optical control capabilities. It can dynamically select the illumination scheme to address different detection challenges, thereby consistently enhancing the contrast between the colloid and the background in various harsh industrial environments, ensuring clear imaging and reliable feature extraction, and greatly improving the system's versatility and detection stability.

[0019] The synchronization control module ensures that the 2D camera and 3D sensor acquire data from the same area at the same instant, fundamentally solving the matching error caused by time asynchrony or incomplete field-of-view overlap of multi-sensor data. The processing module executes image processing and data fusion algorithms, and also directly performs quality judgment based on the fusion results.

[0020] It should be understood that the above general description and the following detailed description are merely exemplary and do not limit the invention. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0022] Figure 1 A schematic diagram of the composition of an optical monitoring system for quality inspection according to an embodiment of the present invention is shown.

[0023] Figure 2 A schematic diagram of the composition of an optical adjustment mechanism according to an embodiment of the present invention is shown.

[0024] Figure 3 A flowchart of an optical monitoring method for quality inspection according to an embodiment of the present invention is shown. Detailed Implementation

[0025] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein; rather, they are provided so that the invention will be thorough and complete, and the concept of the exemplary embodiments will be fully conveyed to those skilled in the art. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted.

[0026] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a full understanding of embodiments of the invention. However, those skilled in the art will recognize that the technical solutions of the invention can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of the invention.

[0027] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0028] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0029] It should be understood that although the terms first, second, third, etc., may be used herein to describe various components, these components should not be limited by these terms. These terms are used to distinguish one component from another. Therefore, the first component discussed below may be referred to as the second component without departing from the teachings of the present invention. As used herein, the term "and / or" includes all combinations of any one and more of the associated listed items.

[0030] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of exemplary embodiments, and the modules or processes in the drawings are not necessarily essential for implementing the present invention, and therefore cannot be used to limit the scope of protection of the present invention.

[0031] To overcome the shortcomings of traditional quality inspection methods, such as insufficient accuracy and poor adaptability under complex working conditions, this invention provides an optical monitoring system and method for quality inspection. By introducing a hardware synchronization mechanism between multimodal illumination and 2D / 3D sensors, comprehensive perception of the adhesive path morphology is achieved.

[0032] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0033] Figure 1 A schematic diagram of an optical monitoring system for quality inspection according to an embodiment of the present invention is shown.

[0034] See Figure 1 The optical monitoring system for quality inspection according to this embodiment includes an imaging module 01, an illumination module 02, a synchronization control module 03, and a processing module 04.

[0035] According to an example embodiment, the imaging module 01 includes at least one camera and a 3D sensor for acquiring image data and three-dimensional topographic data of the dispensing area. The illumination module 02 includes at least two light sources with different incident angles or different polarization states for providing adjustable illumination to the dispensing area to enhance the contrast between the adhesive and the background. The synchronization control module 03 is configured to synchronously trigger the imaging module and coordinate the switching of the light sources in the illumination module. The processing module 04 is connected to the imaging module 01 and the synchronization control module 03, and is configured to perform fusion processing on the synchronously acquired image data and three-dimensional topographic data, and determine the dispensing quality based on the fusion result.

[0036] According to some embodiments, the illumination module 02 includes at least two light sources with different incident angles or different polarization states, such as at least two of ring light, side light and coaxial light. The synchronization control module 03 is also configured to dynamically switch or combine different light sources according to the surface characteristics or colloid type of the product being tested.

[0037] According to some embodiments, the processing module 04 performs fusion processing on the acquired image data and three-dimensional topography data, including: extracting two-dimensional contour information and three-dimensional height information of the adhesive path, and calculating the width, height and cross-sectional area of ​​the adhesive path based on the two-dimensional contour information and the three-dimensional height information.

[0038] The processing module 04 is configured to continuously acquire multiple images of the moving product in the imaging module 01 to obtain a multi-frame continuous image sequence and / or a three-dimensional data sequence; analyze the continuity of the adhesive path trajectory in the sequence; and, based on the continuity analysis results, distinguish between adhesive breakage defects and / or visual pseudo-defects caused by adhesive flow.

[0039] The processing module 04 is further configured to perform positioning and registration of the image data and three-dimensional topography data based on Mark points on the product. The dispensing area is divided into multiple sub-regions, and each sub-region has a preset independent pass / fail threshold for the width and / or height of the adhesive path.

[0040] According to some embodiments, the optical monitoring system for quality inspection further includes an optical adjustment mechanism for aligning the field of view of the camera with that of the 3D sensor in the dispensing area.

[0041] Specifically, see Figure 2 The optical adjustment mechanism 05 can be a multi-degree-of-freedom precision mechanical adjustment frame, serving as the mounting base for the entire vision system. The optical adjustment mechanism 05 includes a base plate 51, a vertical adjustment unit 52, a horizontal adjustment unit 53, and an angle adjustment unit 54.

[0042] The base plate 51 is rigidly fixed to the production line frame or an independent support structure by bolts, providing a stable mounting base for the entire adjustment mechanism. The vertical adjustment unit 52 is mounted on the base plate 51 and includes a high-precision linear guide rail and a graduated manual (or electric) fine-tuning knob. The imaging module 01 is mounted on this unit. By rotating the fine-tuning knob, the imaging module 01 can be precisely moved along a direction perpendicular to the product's movement plane (Z-axis) to adjust the working distance and ensure the field of view covers the entire dispensing path.

[0043] The horizontal adjustment unit 53 is linked to the vertical adjustment unit 52 and integrated on the same adjustment frame, allowing the imaging module 01 to be translated in the horizontal plane (X and Y directions). During initial installation, the horizontal adjustment unit 53 can be used for coarse adjustment so that the center of the field of view of the imaging module 01 is roughly aligned with the adhesive dispensing area on the product.

[0044] The angle adjustment unit 54 is typically implemented as a "one-dimensional or two-dimensional pitch and yaw adjustment platform," installed between the imaging module 01 and the vertical adjustment unit 52. It allows for fine-tuning of the attitude of the imaging module 01, enabling small-angle rotation around the X and Y axes. The angle adjustment unit 54 is the core component for achieving precise overlap between the fields of view of the 2D camera and the 3D sensor.

[0045] In this embodiment, position adjustment is performed using an optical adjustment mechanism. For example, a calibration plate with high-contrast markings (such as a crosshair target) is placed at the dispensing station. First, the imaging module 01 is moved to a suitable working position using the horizontal adjustment unit 53 and the vertical adjustment unit 52; then, the pitch and yaw angles of the imaging module 01 are finely adjusted using the angle adjustment unit 54. The real-time image from the 2D area array camera and the position of the laser line from the 3D sensor (such as a line laser profilometer) on the target are observed. This continues until the center of the 2D camera image and the center line of the laser line from the 3D sensor on the target are completely aligned, and the laser line clearly and completely covers the dispensing path area to be detected in the 2D camera image. At this point, it can be considered that the fields of view of both have achieved precise overlap in the target area. All adjustment units are equipped with locking devices, which are locked after adjustment to ensure that the optical path is stable and does not shift during subsequent use.

[0046] This embodiment, through an optical adjustment mechanism with multiple degrees of freedom (translation and rotation), ensures the physical possibility of spatial alignment between image data and three-dimensional shape data, indirectly guaranteeing the accuracy and effectiveness of the entire fusion detection algorithm, and improving the reliability, adaptability and maintainability of the system in engineering practice.

[0047] The synchronization control module is also configured to dynamically switch or combine different light sources based on the surface characteristics or colloid type of the product being tested. The illumination module includes at least two light sources with different incident angles or different polarization states. For example, to address the challenge of detecting transparent or translucent colloids (such as silicone or UV adhesives), this embodiment employs a "bright field + dark field" hybrid illumination strategy, allowing the illumination module to use coaxial light and side light.

[0048] Coaxial lighting is mainly used to illuminate flat PCB substrates or metal leads. The light is incident perpendicularly and reflected perpendicularly back to the camera, making the background bright (white field). Side lighting is mainly used to highlight the outline of the colloid. Because the surface of the transparent colloid is smooth, the side lighting will be specularly reflected on the colloid surface, and most of the light will not enter the camera. Therefore, the colloid appears as a dark outline in the image.

[0049] The coaxial light and side light are switched on and off by a synchronous control module, and two images are acquired respectively. Background features are extracted from the coaxial light image (for localization), and the edges of the adhesive path are extracted from the side light image (for width measurement). Due to the high contrast between the bright background and the dark outline of the adhesive, the problem of indistinguishable transparent adhesive from the background is solved.

[0050] This embodiment can effectively eliminate interference from highly reflective substrates (such as gold wires and copper foils), while enhancing the edge sharpness of transparent colloids and improving the image signal-to-noise ratio to over 28dB.

[0051] Specifically, the imaging module 01 is fixed to the bracket via an optical adjustment mechanism 05 (such as a precision gimbal and mounting plate), and the field of view of the imaging module 01 is aligned with the dispensing station on the production line. The imaging module 01 includes a high-resolution 2D area array camera (such as a 5-megapixel CMOS industrial camera) and a line laser profilometer (as a 3D sensor). For a line laser sensor, it is only necessary to adjust the laser line to coincide with the center of the camera's field of view; for an area array 3D sensor, it is necessary to adjust the optical axes of both to be parallel or to ensure complete coverage of the field of view.

[0052] The illumination module 02 includes a ring-shaped LED light source and a low-angle strip LED light source, both with different incident angles, mounted around the imaging module 01. The ring light provides uniform front illumination, suitable for most situations; the low-angle light highlights surface texture and edges, suitable for detecting adhesive tracks on highly reflective surfaces.

[0053] The synchronization control module 03 is implemented by a programmable synchronization trigger, which is activated by a photoelectric sensor. The module first selects the corresponding lighting scheme based on the preset product model (e.g., a combination of ring light and low-angle light for black plastic workpieces), and then simultaneously sends trigger signals to the 2D camera, line laser profilometer, and selected light source, ensuring that a synchronous data acquisition is completed during the stable illumination of the light source. By synchronously acquiring high-resolution two-dimensional images and accurate three-dimensional topographic data, the system can simultaneously and at the same location measure the width, height, cross-sectional area, and even volume of the adhesive path, achieving a comprehensive, high-precision, three-dimensional evaluation of dispensing quality. The hard synchronization mechanism provides high-quality, spatiotemporally strictly aligned input data for subsequent processing modules to perform precise "data fusion processing," which is the foundation for ensuring the final detection accuracy. Finally, the processing module 04 performs image processing and analysis.

[0054] This embodiment presents a complete and clearly defined modular system. Multiple functional modules each perform their specific functions while cooperating in a standardized manner, making the system easy to install, calibrate, and maintain. Lighting strategies can be changed via software configuration without requiring hardware replacement; synchronization logic is managed by a dedicated module, reducing system integration complexity and significantly improving the engineering convenience and reliability of the equipment.

[0055] This invention provides an optical monitoring method for quality inspection.

[0056] The optical monitoring system used for quality inspection has multiple preset light source schemes, each corresponding to a product type, such as "UV adhesive scheme," "sealant scheme," or different workpiece surfaces, such as "high reflective metal scheme" or "black acrylic scheme." Each scheme defines the on / off state, brightness, and polarization state of different light sources (such as ring light, coaxial light, and low-angle light) in the lighting module.

[0057] See Figure 3 In S101, the imaging module acquires image information of the dispensing area under the selected illumination mode.

[0058] According to some embodiments, when the product to be tested arrives at the testing station, the corresponding illumination mode instruction is retrieved from the preset library. The synchronous control module sends the illumination mode instruction to the controller of the illumination module, driving the corresponding light source to light up before the imaging module is exposed and enter a stable working state. For example, for transparent UV adhesive, the system automatically selects the "coaxial light + specific polarization angle" mode to suppress background reflection and highlight the outline of the transparent adhesive.

[0059] In S103, the imaging module is triggered by the synchronization control module to synchronously acquire image data and three-dimensional morphology data of the dispensing area.

[0060] According to some embodiments, the synchronization control module is a programmable logic controller (PLC) or a dedicated synchronization trigger. Upon receiving a product arrival signal, it generates a main trigger pulse. Synchronous acquisition execution: This main trigger pulse is simultaneously (or after precise delay compensation) sent to the 2D area scan camera, triggering it to perform a global exposure and acquire a high-resolution 2D grayscale or color image. If the 3D sensor is a line laser profilometer, the pulse triggers the laser to emit and initiates sensor scanning, acquiring 3D point cloud data.

[0061] Each main trigger pulse's initiation timestamp is uniquely bound to the corresponding image data and 3D topography data frame to ensure correct pairing during subsequent processing. Multiple consecutive acquisitions are performed during the product's movement to obtain a time-series data sequence.

[0062] In S105, the processing module fuses the synchronously acquired image data and three-dimensional topography data to obtain fused data containing two-dimensional contour information and three-dimensional height information of the adhesive path.

[0063] According to some embodiments, the processing module identifies at least two pre-made Mark points on the product in the synchronously acquired image data (2D image). Based on the deviation between the identified Mark point positions and the theoretical positions of Mark points in the standard template, a spatial transformation matrix (including translation ΔX, ΔY and rotation Δθ) is calculated.

[0064] The above transformation matrix is ​​applied to the synchronously acquired three-dimensional topography data (3D point cloud data), and the same rotation and translation transformation is performed on the 3D point cloud so that the 2D image and the 3D point cloud data are aligned in the same world coordinate system.

[0065] In the aligned coordinate system, for each pixel belonging to the adhesive path in the 2D image, its height (Z coordinate) value is found in the transformed 3D point cloud based on its image coordinates. In the 2D image, the width is measured as the distance between the two edges perpendicular to the centerline of the adhesive path. The height is directly taken from the Height_3D value of the corresponding position in the fused data, which can be implemented through a coordinate mapping table or interpolation algorithm. Finally, a data pair (X, Y, Width_2D, Height_3D) is generated for each position point of the adhesive path, forming the "fused data".

[0066] In S107, based on the fused data, the two-dimensional contour information and three-dimensional height information are compared with a preset standard to determine whether the dispensing quality is qualified.

[0067] According to some embodiments, the processing module divides the dispensing path into multiple sub-regions (such as straight line segments and endpoints), each with its own preset "acceptable standard value" and "error range". For example, the standard width of a straight line segment is 1.0mm ± 0.1mm, and the height is 0.5mm ± 0.05mm. The actual width and height calculated using the fusion data of the two-dimensional contour information and three-dimensional height information of the dispensing path are compared with the standard range of the corresponding sub-region. If the actual value at any position exceeds the allowable range of the corresponding sub-region, or if the continuity is determined to be a dispensing break, the dispensing quality of the product is deemed unacceptable (NG).

[0068] If a defect is detected, the processing module outputs a final "NG" signal, which may include the specific defect type (such as "insufficient glue in area A" or "insufficient glue at point B") and quantitative data. It then generates a corresponding dispensing process parameter compensation instruction and sends the instruction to the dispensing equipment or production management system to adjust the dispensing process for subsequent products.

[0069] The processing module is also equipped with a time-series analysis-based algorithm unit to distinguish between glue breakage and stringing defects, in order to address scenarios where the product is moving continuously on a conveyor belt.

[0070] Specifically, firstly, image sequence acquisition is performed. The processing module controls the imaging module to operate at a fixed frequency (e.g., hundreds to thousands of frames per second) that matches the production line speed, continuously capturing images of the product dispensing area to obtain an image sequence containing the adhesive path. The sequence consists of multiple consecutive frames of images.

[0071] Extract the adhesive path region, process each frame of the image sequence, including filtering and denoising, and extract the binarized region of the adhesive path using an image segmentation algorithm (e.g., adaptive threshold segmentation), denoted as a binary mask image.

[0072] Continuity analysis is performed based on inter-frame difference or temporal models. According to some embodiments, continuity analysis based on inter-frame difference involves calculating the difference image between two adjacent binary mask images. Specifically, a logical XOR operation is performed on the mask image Mt of frame t and the mask image Mt+1 of frame t+1 to obtain the difference image Dt. The highlighted areas in the difference image Dt reflect the pixel-level changes in the adhesive path region between the two frames.

[0073] The total number of pixels representing changes in the differential image Dt is counted. If the total number is less than a preset dynamic change threshold, the adhesive path segment is considered to change smoothly between frames and has good continuity. If the total number is greater than or equal to the dynamic change threshold, and the position of the changing area is consistent in multiple consecutive frames, it is determined that there may be an adhesive path break at that position.

[0074] Continuity analysis based on a temporal model allows the multi-frame sequence of binary mask images to be input into a pre-trained temporal neural network model for analysis. For example, a recurrent neural network (RNN) or its variants (such as a Long Short-Term Memory network, LSTM) can be used. This model, through learning, can memorize the state evolution of the adhesive path over time.

[0075] The model receives the mask image of the current frame as input and combines it with the context information of previous frames stored in its internal memory to output a probability value indicating whether there is a "break" defect in the adhesive path in the current frame. The processing module 04 compares this probability value with a preset judgment threshold. For example, if the probability value is higher than the first threshold (e.g., 0.9), it is judged as a real "adhesive break" defect; if the probability value is lower than the second threshold (e.g., 0.1), it may be judged as a false defect such as "adhesive stringing" or normal fluid deformation; if it is in the middle range, it can be output as "suspicious" or trigger a more detailed review.

[0076] Based on the results of one or two of the above analysis methods, the processing module 04 generates a final continuity judgment conclusion, such as "continuous" or "fractured", and outputs the conclusion along with the defect location information for subsequent quality judgment and closed-loop control.

[0077] The method in this embodiment involves the processing module 04 receiving synchronously acquired 2D images and 3D data. Two pre-set circular marker points on the product are identified in the 2D image, and the positional offset (ΔX, ΔY) and rotation angle Δθ of the current product relative to the standard template are calculated. Using this transformation matrix, the same coordinate transformation is performed on the 3D point cloud data to align the current product data with the standard template.

[0078] Within the aligned coordinate space, the pixel coordinates of the 2D contour of the 2D image are correlated with the 3D height information of the 3D point cloud. Along the preset dispensing path, the 2D pixel width of the dispensing cross section at each location and the height value from the 3D point cloud are extracted, and then the approximate area of ​​each cross section is calculated. Finally, the total volume is estimated by integrating along the path.

[0079] Further comparison with the preset standards reveals that, assuming a straight section of sealing adhesive on the product, the customer's acceptable standards (i.e., preset standards) are: "Standard width (W_std): 2.0 mm, width tolerance (ΔW): ±0.2 mm; standard height (H_std): 1.0 mm, height tolerance (ΔH): ±0.1 mm."

[0080] The processing module inspects the current product and performs data fusion and calculation at a specific location (section A) of the straight line segment: Contour information extracted from the 2D image is used to calculate the actual pixel width of the adhesive path at that section, which is then converted to physical dimensions through camera calibration, resulting in an actual width of (W_act) = 1.85 mm. Height information extracted from the 3D point cloud is used to read the actual height of the section at the corresponding location from the 3D point cloud data, which is (H_act) = 1.12 mm.

[0081] The processing module compares the actual calculated value with the preset standard value: Width difference calculation: |W_act - W_std| = |1.85 - 2.0| = 0.15 mm Threshold comparison: 0.15 mm < ΔW (0.2 mm) Result: The width difference is within the allowable error range.

[0082] Height difference calculation: |H_act - H_std| = |1.12 - 1.0| = 0.12 mm Threshold comparison: 0.12 mm > ΔH (0.1 mm) Result: The height difference exceeds the allowable error range.

[0083] Based on comprehensive assessment, in this example, the actual height of section A (1.12 mm) exceeds the upper limit of the acceptable range (1.1 mm). Although the width is acceptable, the height is not. Therefore, the processing module determines that the product has an "over-adhesion" (or "height out of tolerance") defect at this location, and ultimately outputs that the dispensing quality of the product is NG (unacceptable), and records the defect type and location (e.g., "straight segment section A, height exceeds the upper limit").

[0084] In another embodiment of this method, temporal continuity analysis is performed on a high-speed production line. For example, the processing module controls the imaging module to continuously acquire 5 frames of data at a high frame rate. The position of the glue path centerline in these 5 frames of data is analyzed, and the continuity of the glue path trajectory between adjacent frames is calculated (e.g., whether the distance between the center points of adjacent frames exceeds a threshold, or a simple sliding window algorithm is used to determine whether the trajectory is interrupted). If trajectory interruption is detected in multiple consecutive frames (e.g., more than 3 frames), it is determined as "glue breakage"; if only one or two frames show a suspected interruption, but subsequent frames restore continuity, it may be a "stringing" artifact, which can be determined as acceptable or a warning.

[0085] The calculated adhesive path width and height are compared with the preset standard values ​​and tolerance zones. Simultaneously, it is checked whether the adhesive path covers all specified path areas. If any of the width, height, or coverage exceeds the standard, or if the timing analysis determines that the adhesive is broken, an NG (non-conforming) signal is output along with the specific defect type and location.

[0086] According to some embodiments, the processing module can also be configured with closed-loop feedback. If multiple products are detected to have "insufficient glue" defects at a specific location, the processing module can generate an instruction to increase the opening time of the dispensing valve at that location by 5%, and send it to the dispensing controller via the industrial network to achieve automatic fine-tuning of process parameters.

[0087] The optical monitoring method and system of the present invention can switch the illumination module according to the actual application scenario for detecting the coating of transparent UV adhesive on glass. It adopts a combination of coaxial light and polarized light at a specific angle to eliminate glass reflection and make the transparent adhesive path clear image.

[0088] In terms of judgment logic, the processing module divides a complex sealant path into several sub-regions, including straight segments, arc segments, and right-angle inflection point segments. Stricter width tolerances (e.g., ±0.08mm) are set for straight segments, while looser tolerances (e.g., ±0.15mm) are set for inflection point segments to accommodate the natural accumulation of adhesive at corners. This region-based judgment method further improves the accuracy and practicality of the detection. The processing module not only performs complex image processing and data fusion algorithms but also directly performs quality judgments based on the fusion results.

[0089] The design of this invention enables the system to automatically complete the entire process from "image acquisition" to "quality judgment," outputting instructions that can be directly used for production line sorting or process adjustment. This provides a complete hardware and logic foundation for realizing online, real-time, and automated quality inspection and subsequent closed-loop feedback control of process parameters (such as automatically fine-tuning the dispensing amount based on the inspection results).

[0090] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0091] Those skilled in the art will clearly understand that the technical solutions of the present invention can be implemented by means of software and / or hardware. In this specification, "unit" and "module" refer to software and / or hardware capable of independently performing or cooperating with other components to perform a specific function, wherein the hardware may be, for example, a field dynamic gate array, integrated circuit, etc.

[0092] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0093] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0094] In the several embodiments provided by this invention, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some service interface; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0095] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0096] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0097] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention.

[0098] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0099] Exemplary embodiments of the present invention have been specifically shown and described above. It should be understood that the present invention is not limited to the detailed structures, arrangements, or implementations described herein; rather, the present invention is intended to cover various modifications and equivalent arrangements contained within the spirit and scope of the appended provisions.

Claims

1. An optical monitoring system for quality inspection, characterized in that, include: An imaging module, comprising at least one camera and a 3D sensor, for acquiring image data and three-dimensional topography data of the dispensing area; An illumination module comprising at least two light sources with different incident angles or different polarization states, for providing adjustable illumination to the dispensing area to enhance the contrast between the adhesive and the background; A synchronization control module is configured to synchronously trigger the imaging module and coordinate the light source switching of the illumination module; The processing module is connected to the imaging module and the synchronization control module, and is configured to perform fusion processing on the acquired image data and three-dimensional topography data, and judge the dispensing quality based on the fusion result.

2. The system according to claim 1, characterized in that, The lighting module includes at least two of the following: ring light, side light, and coaxial light; The synchronization control module is also configured to dynamically switch or combine different light sources based on the surface characteristics or colloid type of the product being tested.

3. The system according to claim 1, characterized in that, The camera may be a 2D camera or a 3D camera; The 3D sensor includes a line laser profilometer, a structured light projector, or a binocular stereo vision camera.

4. The system according to claim 1, characterized in that, It also includes an optical adjustment mechanism for making the field of view of the camera and the 3D sensor overlap in the dispensing area.

5. The system according to claim 1, characterized in that, The processing module performs fusion processing on the synchronously acquired image data and three-dimensional topography data, including: extracting the two-dimensional contour information and the three-dimensional height information of the adhesive path, and calculating the width, height and cross-sectional area of ​​the adhesive path based on the two-dimensional contour information and the three-dimensional height information.

6. The system according to claim 1, characterized in that, The processing module is configured as follows: The imaging module continuously acquires multiple images of the moving product to obtain a series of consecutive image frames and / or a three-dimensional data sequence. Analyze the continuity of the adhesive path trajectory in the sequence; Based on the continuity analysis results, distinguish between glue breakage defects and / or visual pseudo-defects caused by glue flow.

7. The system according to claim 1, characterized in that, The processing module is configured to perform localization and registration of the image data and three-dimensional shape data based on the Mark points on the product. The processing module is further configured to divide the dispensing area into multiple sub-regions, and to preset independent qualification thresholds for the width and / or height of the adhesive path for each sub-region.

8. An optical monitoring method for quality inspection, characterized in that, Applied to the system as described in any one of claims 1-7, the method comprises: The imaging module acquires image information of the dispensing area under the selected lighting mode; The synchronous control module triggers the imaging module to synchronously acquire image data and three-dimensional morphological data of the dispensing area; The processing module fuses the synchronously acquired image data and three-dimensional topography data to obtain fused data containing two-dimensional contour information and three-dimensional height information of the adhesive path. Based on the fused data, the two-dimensional contour information and three-dimensional height information are compared with preset standards to determine whether the dispensing quality is qualified.

9. The method according to claim 8, characterized in that, The step of triggering the imaging module through the synchronization control module to synchronously acquire image data and three-dimensional morphology data of the dispensing area includes: Multiple consecutive data acquisitions are performed during the product's movement to obtain a time-series data sequence; The inter-frame continuity of the adhesive path trajectory in the time-series data sequence is analyzed, and based on the continuity analysis results, it is determined whether there is an adhesive breakage defect in the adhesive path.

10. The method according to claim 9, characterized in that, Also includes: If a defect is identified, a corresponding dispensing process parameter compensation instruction is generated. The process parameter compensation command is sent to the dispensing equipment or production management system to adjust the dispensing process of subsequent products.

Images