A kind of binocular camera metal casting surface defect measuring device based on machine vision

Abstract

The utility model discloses a kind of binocular camera metal casting surface defect measuring devices based on machine vision, it is related to industrial detection technical field.The device includes detection platform, two light sources symmetrically arranged above platform, binocular camera located above platform, image acquisition and processing workstation and display screen.The image acquisition and processing workstation are equipped with gray scale conversion module, filter module, photometric correction module, stereo matching module, three-dimensional coordinate calculation module, surface analysis module and defect quantification module.The utility model obtains high-precision three-dimensional point cloud data of metal casting surface by binocular stereo vision technology, can accurately identify and quantify the three-dimensional size of crack, pore, slag inclusion, shrinkage and the like, such as depth, volume, effectively overcome the limitations of artificial detection and traditional two-dimensional vision detection, with the advantages of high detection precision, high efficiency, high degree of automation.

Description

Technical Field

[0001] This utility model relates to the field of machine vision inspection technology, specifically to a binocular camera-based device for measuring surface defects in metal castings based on machine vision. Background Technology

[0002] This section provides only background information related to this application to enable those skilled in the art to understand this application more thoroughly and accurately, and it is not necessarily prior art.

[0003] Metal castings are indispensable basic components in key industrial fields such as aerospace, automotive manufacturing, and heavy machinery. Their surface quality directly determines the performance, reliability, and service life of the final product. Common surface defects such as cracks, porosity, inclusions, and shrinkage cavities are not only stress concentration points that significantly reduce the fatigue strength of parts, but also potential safety hazards. Therefore, accurate and efficient surface defect detection of castings during the production process is crucial.

[0004] Currently, traditional surface defect detection methods mainly rely on manual visual inspection or the use of contact measuring tools (such as calipers and probes). Manual inspection is highly subjective and easily affected by factors such as the inspector's experience, fatigue, and emotional state, leading to high rates of missed and false detections, low efficiency, and difficulty in digitizing and archiving results, thus failing to meet the demands of modern large-scale, standardized industrial production. Contact measurement methods are not only inefficient but also risk causing secondary scratches to the surface of precision parts, and cannot accurately obtain the three-dimensional geometric information of defects (such as depth and indentation volume).

[0005] In recent years, machine vision technology has been widely used in industrial inspection due to its non-contact, high efficiency, and high precision. However, traditional two-dimensional vision systems based on monocular cameras are unable to overcome interference such as metal surface reflection and uneven lighting, and are inherently unable to acquire high-precision three-dimensional depth information, thus having inherent limitations when measuring key three-dimensional dimensions such as the depth and volume of defects. Utility Model Content

[0006] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a binocular camera-based metal casting surface defect measurement device based on machine vision, so as to solve the shortcomings of traditional manual inspection and two-dimensional visual inspection in terms of efficiency, accuracy and three-dimensional quantization.

[0007] The technical solution adopted by this utility model to solve its technical problem is: a binocular camera-based metal casting surface defect measurement device based on machine vision, comprising: a detection platform, a first light source, a second light source, a binocular camera, an image acquisition and processing workstation, and a display screen; the detection platform is used to place the metal casting to be inspected; the first light source and the second light source are arranged axially symmetrically on both sides above the detection platform to provide uniform illumination for the surface of the metal casting.

[0008] The binocular camera is positioned above the inspection platform with its optical axis aligned with the center of the platform, and is used to synchronously acquire binocular images of the surface of the illuminated metal casting. The image acquisition and processing workstation is communicatively connected to the binocular camera and is used to receive and process the binocular images and calculate the three-dimensional information of the defects. The display screen is connected to the image acquisition and processing workstation and is used to display the images and inspection results.

[0009] According to an example of this utility model, the image acquisition and processing workstation integrates multiple image processing modules connected in sequence, including: a grayscale conversion module for converting an image into a grayscale image, a filtering module for suppressing image noise, a photometric correction module for correcting image distortion, a stereo matching module for calculating pixel parallax, a three-dimensional coordinate calculation module for reconstructing a three-dimensional point cloud of the surface, a surface analysis module for segmenting defect regions, and a defect quantization module for calculating defect size.

[0010] According to one example of the present invention, the binocular camera includes two axially symmetrically arranged industrial cameras, each equipped with a fixed-focus industrial lens suitable for industrial inspection.

[0011] According to one example of the present invention, a camera translation device is also included, on which the binocular camera is fixedly mounted. The camera translation device is used to drive the binocular camera to move in three-dimensional space to adapt to metal castings of different sizes.

[0012] According to one example of the present invention, a rigid gantry support structure is also included, the gantry support structure including a crossbeam spanning above the detection platform, and the binocular camera is mounted on the crossbeam via a camera translation device.

[0013] According to one example of the present invention, the camera translation device includes mutually orthogonal X-axis, Y-axis and Z-axis linear slide modules, and the binocular camera is fixedly mounted on the Z-axis linear slide module.

[0014] According to one example of the present invention, the first light source and the second light source are LED strip light sources with adjustable brightness in multiple segments, and the illumination angle of the LED strip light source is adjustable.

[0015] According to one example of the present invention, the first light source and the second light source are ring-shaped LED light sources with adjustable brightness and color temperature. The ring-shaped LED light sources are coaxially arranged around the industrial lens of the binocular camera to eliminate reflections on the metal surface and provide shadowless illumination.

[0016] The beneficial effects of this utility model are:

[0017] (1) Using binocular stereo vision technology, high-precision three-dimensional point cloud data of the surface can be obtained without contacting the workpiece, avoiding secondary damage, and accurately measuring the depth, volume and other three-dimensional indicators that are difficult to obtain by traditional methods.

[0018] (2) The design of the camera translation device and adjustable light source enables the device to quickly adapt to workpieces of different sizes and different detection requirements, making it highly versatile.

[0019] (3) The entire testing process is completed automatically, which greatly improves the testing efficiency. It is suitable for online real-time testing on the production line and overcomes the problems of low efficiency and poor consistency of manual testing.

[0020] Additional aspects and advantages of this invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this utility model, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this utility model. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of the overall device structure of this utility model.

[0023] Figure 2 This is a schematic block diagram of the image processing module within the image acquisition and processing workstation of this utility model.

[0024] The attached figures are labeled as follows:

[0025] 1. First light source; 2. Metal casting; 3. Binocular camera; 4. Second light source; 5. Detection platform; 6. Image acquisition and processing workstation; 7. Display screen. Detailed Implementation

[0026] The embodiments of this utility model are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this utility model, and should not be construed as limiting this utility model.

[0027]

Example

[0028] Please see Figure 1 , Figure 2 As shown, this utility model provides a machine vision-based binocular camera surface defect measurement device for metal castings, including: a detection platform 5, a first light source 1, a second light source 4, a binocular camera 3, an image acquisition and processing workstation 6, and a display screen 7. The detection platform 5 supports the metal casting 2 to be inspected. The first light source 1 and the second light source 4 are arranged symmetrically on both sides above the detection platform 5 to provide a uniform and controllable lighting environment for the surface of the metal casting 2, reducing reflections and highlighting defect features. The binocular camera 3 is fixed above the detection platform 5 and is used to simultaneously acquire two two-dimensional images of the illuminated surface of the metal casting 2. The image acquisition and processing workstation 6 is communicatively connected to the binocular camera 3 and is used to receive and process the binocular images and calculate the three-dimensional information of the defects based on a stereo vision algorithm. The display screen 7 is connected to the image acquisition and processing workstation 6 and is used to display the acquired images, processing process, three-dimensional model, and final defect detection report in real time.

[0029] The image acquisition and processing workstation 6 integrates multiple image processing modules connected in sequence, including: a grayscale conversion module for converting images to grayscale images, a filtering module for suppressing image noise, a photometric correction module for correcting image distortion, a stereo matching module for calculating pixel parallax, a three-dimensional coordinate calculation module for reconstructing the three-dimensional point cloud of the surface, a surface analysis module for segmenting defect regions, and a defect quantization module for calculating defect size.

[0030] Through a modular image processing workflow, fully automated processing from raw images to defect quantification data is achieved. Grayscale conversion simplifies processing complexity; the filtering module effectively improves the signal-to-noise ratio; photometric correction ensures the geometric authenticity of the image, laying the foundation for accurate 3D measurement; stereo matching and 3D coordinate calculation are the core of acquiring high-precision point cloud data; and the surface analysis and defect quantification modules directly output quantitative indicators of defects, achieving objectivity and accuracy in detection.

[0031] The binocular camera 3 comprises two symmetrically arranged industrial cameras equipped with fixed-focus industrial lenses suitable for industrial inspection. The use of high-performance industrial cameras and high-quality fixed-focus lenses ensures high resolution, low noise, and low optical distortion in the acquired images, enabling high-precision stereo matching and 3D reconstruction, and guaranteeing the accuracy and reliability of measurements.

[0032] In this embodiment, a camera translation device is also included. The binocular camera 3 is fixedly mounted on the camera translation device, which drives the binocular camera 3 to move in three-dimensional space to adapt to metal castings 2 of different sizes. This device greatly improves the flexibility of the system. By adjusting the position of the camera, it can adapt to the inspection needs of workpieces of various sizes, from small precision castings to large structural parts, thus expanding the application range of the equipment.

[0033] In this embodiment, a rigid gantry support structure is also included. The gantry support structure includes a crossbeam spanning above the detection platform 5, and the binocular camera 3 is mounted on the crossbeam via a camera translation device. The gantry support structure provides extremely high rigidity and stability, effectively suppressing external vibrations and preventing structural deformation, providing a stable mounting reference for the binocular camera 3, and ensuring the long-term stability of the camera's optical parameters and the repeatability of the measurement results.

[0034] Furthermore, the camera translation device includes mutually orthogonal X-axis, Y-axis, and Z-axis linear slide modules, with the binocular camera 3 fixedly mounted on the Z-axis linear slide module. This three-axis orthogonal precision mechanical structure allows the operator to make precise and flexible adjustments to the camera position. Z-axis movement is used to adjust the field of view size and resolution, while X and Y-axis movement is used to precisely align with specific areas of the workpiece under test, achieving precise positioning and meeting the inspection needs of complex workpieces.

[0035] As one feasible design, the first light source 1 and the second light source 4 are LED strip light sources with adjustable brightness in multiple stages, and the illumination angle of the LED strip light sources is adjustable. The strip light sources with adjustable brightness and angle provide flexible lighting solutions. For defects of different materials (high reflectivity, matte) and different shapes, the contrast between defects and background can be maximized and unnecessary reflections can be suppressed by optimizing the illumination angle and intensity, thereby obtaining the original image most conducive to image processing and defect recognition.

[0036] As an alternative feasible design, the first light source 1 and the second light source 4 are adjustable brightness and color temperature ring-shaped LED light sources. These ring-shaped LED light sources are coaxially arranged around the industrial lens of the binocular camera 3 to eliminate reflections on metal surfaces and provide shadowless illumination. Coaxial illumination is an effective means of eliminating specular reflections on metal surfaces. The ring-shaped light source provides uniform shadowless illumination, clearly highlighting minute undulations and unevenness (such as scratches and dents) on the object's surface. It is particularly suitable for detecting defects on metal surfaces with complex curvature or high reflectivity, significantly improving the system's signal-to-noise ratio and defect detection capability.

[0037] The steps for using this utility model device are as follows:

[0038] S1. System Calibration: The binocular camera 3 is calibrated using a high-precision checkerboard calibration board to obtain the camera's intrinsic parameters (focal length, principal point, distortion coefficient) and extrinsic parameters (rotation matrix and translation vector) and to establish an accurate binocular vision model.

[0039] S2. Workpiece clamping and adjustment: Place the metal casting 2 to be inspected on the inspection platform 5. Adjust the X, Y, and Z positions of the binocular camera 3 by operating the camera translation device to ensure that its field of view completely covers the area to be inspected and that it is clearly focused. Adjust the brightness, color temperature, and illumination angle of the light source (1, 4) according to the surface characteristics of the workpiece to obtain an image with optimal contrast and no glare interference.

[0040] S3. Image Acquisition: After all parameters are set, the acquisition is triggered through the software interface of the image acquisition and processing workstation 6. The binocular camera 3 simultaneously captures two images and transmits them to the workstation.

[0041] S4. Image Processing and 3D Reconstruction: The software within workstation 6 automatically executes the processing workflow.

[0042] S41. Grayscale Conversion and Filtering: Convert the image to grayscale and perform noise reduction preprocessing.

[0043] S42. Photometric Correction: Corrects lens distortion using calibration parameters.

[0044] S43. Stereo Matching: Calculate the disparity map of two corrected images.

[0045] S44. 3D coordinate calculation: Based on parallax and camera parameters, a high-density 3D point cloud of the workpiece surface is reconstructed.

[0046] S5. Defect Analysis and Quantification:

[0047] S51. Surface Analysis: Mesh and smooth the 3D point cloud, and identify and locate defect areas through normal vector analysis or height threshold segmentation algorithm.

[0048] S52. Defect Quantification: For each segmented defect, calculate its quantitative parameters such as area, maximum depth, average depth, and volume.

[0049] S6. Result Output and Display: The final inspection results, including defect location, type, size data, 3D rendering, etc., are displayed on the display screen 7 in real time, and a detailed inspection report can be generated for archiving or output.

[0050] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this utility model, "a plurality of" means two or more, unless otherwise explicitly specified.

[0051] In this utility model, unless otherwise explicitly specified and limited, the terms "installation," "connection," "joining," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this utility model according to the specific circumstances.

[0052] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

[0053] For those skilled in the art, various changes and modifications will undoubtedly be apparent after reading the above description. Therefore, the appended claims should be considered as covering all changes and modifications that encompass the true intent and scope of this utility model. Any and all equivalent scope and content within the scope of the claims should be considered as still falling within the intent and scope of this utility model.

Claims

1. A binocular camera-based device for measuring surface defects in metal castings based on machine vision, characterized in that, include: The detection platform (5), the first light source (1), the second light source (4), the binocular camera (3), the image acquisition and processing workstation (6), and the display screen (7) are all included. The testing platform (5) is used to place the metal casting (2) to be tested; The first light source (1) and the second light source (4) are arranged on both sides above the detection platform (5) in an axially symmetrical manner to provide uniform illumination to the surface of the metal casting (2); The binocular camera (3) is set above the detection platform (5), with its optical axis facing the center of the detection platform (5), and is used to synchronously acquire binocular images of the surface of the illuminated metal casting (2); The image acquisition and processing workstation (6) is communicatively connected to the binocular camera (3) and is used to receive and process the binocular images and calculate the three-dimensional information of the defects; The display screen (7) is connected to the image acquisition and processing workstation (6) and is used to display images and detection results.

2. The binocular camera-based device for measuring surface defects in metal castings based on machine vision according to claim 1, characterized in that, The image acquisition and processing workstation (6) integrates multiple image processing modules connected in sequence, including: a grayscale conversion module for converting images into grayscale images, a filtering module for suppressing image noise, a photometric correction module for correcting image distortion, a stereo matching module for calculating pixel parallax, a three-dimensional coordinate calculation module for reconstructing surface three-dimensional point clouds, a surface analysis module for segmenting defect regions, and a defect quantization module for calculating defect size.

3. The binocular camera-based device for measuring surface defects in metal castings based on machine vision according to claim 1, characterized in that, The binocular camera (3) includes two axially symmetrically arranged industrial cameras, each equipped with a fixed-focus industrial lens suitable for industrial inspection.

4. The binocular camera-based device for measuring surface defects in metal castings based on machine vision according to claim 3, characterized in that, It also includes a camera translation device, on which the binocular camera (3) is fixedly mounted. The camera translation device is used to drive the binocular camera (3) to move in three-dimensional space to adapt to metal castings of different sizes.

5. The binocular camera-based device for measuring surface defects in metal castings based on machine vision according to claim 4, characterized in that, It also includes a rigid gantry support structure, which includes a crossbeam spanning above the detection platform (5), and the binocular camera (3) is mounted on the crossbeam via a camera translation device.

6. The binocular camera-based device for measuring surface defects in metal castings based on machine vision according to claim 5, characterized in that, The camera translation device includes mutually orthogonal X-axis, Y-axis and Z-axis linear slide modules, and the binocular camera (3) is fixedly installed on the Z-axis linear slide module.

7. The binocular camera-based device for measuring surface defects in metal castings based on machine vision according to claim 1, characterized in that, The first light source (1) and the second light source (4) are LED strip light sources with adjustable brightness in multiple segments, and the illumination angle of the LED strip light source is adjustable.

8. The binocular camera-based device for measuring surface defects in metal castings based on machine vision according to claim 1, characterized in that, The first light source (1) and the second light source (4) are ring-shaped LED light sources with adjustable brightness and color temperature. The ring-shaped LED light sources are coaxially arranged around the industrial lens of the binocular camera (3) to eliminate reflections on the metal surface and provide shadowless illumination.

Images