Multi-angle image acquisition device for complex object surfaces based on multi-camera array

Abstract

This utility model relates to a multi-camera array-based multi-angle image acquisition device for complex object surfaces, comprising a support frame, a multi-camera imaging system, an illumination device, a laser calibration device, and a control system. The multi-camera imaging system includes a base plate, multiple cameras rigidly arranged in an array on the base plate, an orientation control pan-tilt unit, and cameras mounted on the pan-tilt unit. Both the illumination device and the laser calibration device are mounted on the base plate. The illumination device provides uniform illumination to the cameras, and the laser calibration device projects a laser calibration grid covering the target object surface. The control system includes a central processing unit, a comprehensive controller, and a memory, used to drive the cameras to adjust their orientation angles to the pan-tilt unit and to control the cameras to acquire images of the target object surface. This utility model enables comprehensive and flexible high-quality multi-angle and multi-directional imaging of complex object surfaces, effectively reducing image distortion and shadows, and significantly improving the efficiency and accuracy of image acquisition for complex object surfaces.

Description

Technical Field

[0001] This utility model relates to the field of optical imaging technology for complex object surfaces, specifically to a multi-angle image acquisition device for complex object surfaces based on a multi-camera array. Background Technology

[0002] In industrial production, visual inspection of objects with complex surface structures (such as automobile wheel hubs, precision molds, and complex curved surface workpieces) is a crucial step in ensuring product quality. Traditional manual visual inspection methods are inefficient, labor-intensive, and prone to missed inspections and misjudgments due to fatigue. Furthermore, the inspection results are highly subjective and inconsistent.

[0003] To improve the efficiency of appearance quality inspection, some automated optical inspection devices have emerged in the existing technology. These devices typically use one or more cameras at fixed angles to photograph objects, and then analyze defects through image processing algorithms. However, such devices have significant shortcomings: First, because the camera angle is fixed and singular, it is difficult to fully cover all parts of the complex object surface, especially occluded areas such as recesses, edges, and curved backs, which can easily create blind spots. Second, the fixed viewing angle is difficult to adaptively adjust according to the object's shape, resulting in potentially severe perspective distortion or shadows in the captured images, which complicates subsequent image analysis and defect identification. Finally, although some solutions use multiple cameras, the camera poses are fixed, lacking flexibility and unable to optimize shooting for objects of different models and sizes, resulting in poor versatility.

[0004] Therefore, there is an urgent need in this field for an acquisition device that can flexibly, comprehensively, and from multiple angles acquire images of the surface of complex objects. Utility Model Content

[0005] The purpose of this invention is to overcome the shortcomings of existing image acquisition devices and provide a multi-angle image acquisition device for complex object surfaces based on a multi-camera array that can comprehensively and flexibly capture high-quality images of complex object surfaces from multiple angles and directions, thereby improving the efficiency and accuracy of appearance quality inspection.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A multi-camera array-based multi-angle image acquisition device for complex object surfaces includes:

[0008] Supporting framework;

[0009] A multi-camera imaging system mounted on the support frame includes a base plate, a camera orientation control gimbal, and a camera for acquiring images of the surface of a target object. Each camera orientation control gimbal is connected to one camera, and multiple camera orientation control gimbals are evenly distributed in an array and rigidly connected to the base plate.

[0010] An illumination device fixed to the base plate is used to provide supplementary light for the camera to capture images. The illumination device includes at least one elongated light source, which is arranged around the camera array.

[0011] A laser calibration device fixed to the base plate is used to project a laser calibration grid covering the surface of the target object;

[0012] Control system, including:

[0013] The central processing unit is communicatively connected to the integrated controller, the memory, and each of the cameras. It is used to trigger the cameras to acquire images, receive and process the grid images acquired by each of the cameras when the laser calibration device is turned on, output motion control commands to the integrated controller, and receive the images of the target object surface acquired by each of the cameras when the laser calibration device is turned off.

[0014] The integrated controller is communicatively connected to the lighting device, the laser calibration device, and each of the camera orientation control gimbals, and is used to control the opening and closing of the lighting device and the laser calibration device, and to drive the camera orientation control gimbals to move according to the motion control commands;

[0015] A memory communicatively connected to the central processing unit is used to store images of the target object surface captured by the camera.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] (1) This utility model rigidly integrates multiple cameras and independently controllable camera orientation control gimbal in an array on the base plate to form a multi-view synchronous acquisition system, which can simultaneously capture the surface of complex objects (such as car wheel hubs) from different directions and angles, and realizes the acquisition of complex object surfaces without blind spots, full coverage and high efficiency.

[0018] (2) This utility model uses the grid projected by the laser calibration device as a reference. The control system can adaptively calculate and adjust the orientation of each camera toward the control gimbal so that the optical axis of the camera is as perpendicular as possible to the local curved surface being measured. At the same time, combined with the uniform illumination arranged around, it fundamentally reduces the generation of image distortion and local shadows, significantly improves image quality, and provides a high-quality, low-error image source for subsequent high-precision appearance quality inspection.

[0019] (3) This utility model adopts a multi-level rigid connection design from the base plate to the camera-oriented control gimbal, which ensures that the entire camera array remains stable during movement (such as being driven by a robotic arm) and operation, avoids image blurring caused by vibration, and ensures the consistency and reliability of the acquired data, making it very suitable for industrial applications. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the structure of the device of this utility model;

[0021] Figure 2 This is a schematic diagram of the control system.

[0022] Explanation of reference numerals in the attached drawings: 1. Support frame; 2. Base plate; 3. Camera orientation control gimbal; 4. Camera; 5. Illumination device; 6. Laser calibration device; 7. Control system; 7-1. Central processing unit; 7-2. Integrated controller; 7-3. Memory. Detailed Implementation

[0023] The technical solution of this utility model will now be described in detail with reference to the accompanying drawings and preferred embodiments.

[0024] This utility model discloses a multi-camera array-based multi-angle image acquisition device for complex object surfaces, which mainly includes a support frame 1, a multi-camera imaging system, an illumination device 5, a laser calibration device 6, and a control system 7, as shown in the figure. Figure 1 As shown.

[0025] The support frame 1 provides support and stability for the entire data acquisition device. It has a robust structure and a stable base (such as a weighted metal chassis) can be installed at the bottom to ensure the stability of the data acquisition device during operation. Optionally, the bottom of the support frame 1 can be fixed to the robotic arm using bolts or similar means, allowing it to move with the robotic arm.

[0026] The multi-camera imaging system includes a base plate 2, a camera orientation control gimbal 3, and a camera 4, where the camera 4 is used to acquire images of the target object's surface. Multiple camera orientation control gimbals 3 are arranged in an array, uniformly distributed and rigidly connected to the base plate 2. For example, nine camera orientation control gimbals 3 are evenly distributed in a 3×3 array on the base plate 2, forming a tightly integrated structure. The base plate 2 is rigidly connected to the camera orientation control gimbals 3, the illumination device 5, and the laser calibration device 6, ensuring the stability of the multi-connection and allowing the entire acquisition device to remain stable even when moved by a robotic arm or other actuators.

[0027] Each camera 4 is rigidly connected to a separate camera orientation control gimbal 3, and the assembly is rigidly connected to the base plate 2. This multi-level rigid connection design further enhances the stability and overall integrity of the device. The camera orientation control gimbal 3 has the ability to rotate freely within a certain range, i.e., it has pitch and pan functions. Therefore, the camera orientation control gimbal 3 is controlled by the control system 7 to achieve precise adjustment of the orientation of the camera 4. This design allows the camera 4 to flexibly acquire image information from different angles, greatly improving the flexibility and comprehensiveness of image acquisition of complex object surfaces. Optionally, the camera orientation control gimbal 3 in this invention is a small gimbal with two degrees of freedom: pitch and pan. It includes a bracket, a motor, and an angle sensor. The camera 4 is fixedly mounted on the bracket. The motor drives the bracket to perform pitch and / or pan movements under the control of the control system 7. The angle sensor detects the angle of rotation of the motor and feeds the angle value back to the control system 7. Preferably, the camera 4 is a camera with an autofocus function, so that the camera 4 can use its own focus adjustment function to ensure the clarity of the image when shooting. Meanwhile, the focal length of camera 4 can also be set in real time by control system 7.

[0028] The lighting device 5 is fixed to the base plate 2, and its function is to provide supplementary lighting for the camera 4 when acquiring images. During the shooting process of the camera 4, insufficient light can cause shadows in local areas, resulting in unclear images. The lighting device 5 effectively avoids this problem. The lighting device 5 in this invention includes at least one elongated light source, which is arranged around the camera array to provide uniform illumination for the camera 4. Optionally, four elongated light sources are arranged around the camera array to further improve the uniformity of illumination and enhance image quality. Optionally, the elongated light sources use the visible light band, with an emission wavelength of 380-780nm.

[0029] The laser calibration device 6 is also fixed on the base plate 2 and is used to project a laser calibration grid covering the surface of the target object. The laser calibration device 6 is mainly composed of a laser capable of emitting the laser calibration grid. For example, this laser can be implemented using a structured light generator, with a power range of 10-100mW. The specific power value can be set according to the shooting distance of the camera 4. The laser calibration device 6 can project a laser calibration grid of size M×N (e.g., 10×10). By adjusting the base plate 2 or the target object, the laser calibration grid can completely cover the surface of the target object. The camera 4 mounted on the base plate 2 acquires a grid image including the surface of the target object and the grid, and sends the grid image to the control system 7. The control system 7 calculates the deformation of the grid and feeds back the corresponding motion control command to the camera orientation control gimbal 3 of each camera 4, realizing the adaptive adjustment of the shooting angle of each camera 4. This laser calibration grid plays an important role in the camera calibration process. The control system 7 uses the laser calibration grid as a reference and can realize the calibration conversion from curved surface (complex object surface) to plane by calculating the curved laser grid lines.

[0030] The control system 7 mainly consists of a central processing unit 7-1, an integrated controller 7-2, and a memory 7-3, such as... Figure 2As shown in the diagram. The central processing unit (CPU) 7-1 is communicatively connected to the integrated controller 7-2, the memory 7-3, and each camera 4. The integrated controller 7-2 is also communicatively connected to the camera orientation control gimbal 3, the lighting device 5, and the laser calibration device 6. The CPU 7-1 controls the cameras 4 to acquire images, receives and processes the grid images acquired by each camera 4 when the laser calibration device 6 is on, processes the grid images according to its built-in image processing algorithm, and generates and outputs corresponding motion control commands to the integrated controller 7-2. Simultaneously, the CPU 7-1 also receives images of the target object surface acquired by each camera 4 when the laser calibration device 6 is off. The integrated controller 7-2 controls the on / off states of the lighting device 5 and the laser calibration device 6, and feeds back their states to the CPU 7-1. It also drives each camera to perform corresponding pitch and / or sweep movements according to the motion control commands sent by the CPU 7-1. The memory 7-3 stores the images of the target object surface acquired by the cameras 4 under the control of the CPU 7-1. The image processing algorithm built into the CPU 7-1 refers to the calibration transformation from a curved surface (complex object surface) to a plane by calculating the curved laser grid lines. This process is implemented by the CPU 7-1 executing the calibration calculation program. For those skilled in the art, the algorithms that can achieve this calculation function are known and diverse. For example, the coordinates of the laser grid are extracted from the acquired image, combined with the camera's intrinsic and extrinsic parameters, and the three-dimensional point cloud and normal vector of the object surface are calculated using the principles of stereo vision or structured light 3D reconstruction. Then, the angle adjustment required to make the camera's optical axis perpendicular to the local surface is solved. Based on this angle adjustment, the angle at which the camera tilts or pans towards the control gimbal 3 is determined, and the corresponding motion control commands are generated. Optionally, the integrated controller 7-2 adopts a programmable logic controller or an embedded microcontroller.

[0031] Optionally, the integrated controller 7-2 employs a closed-loop control algorithm, such as a PID control algorithm, to adjust the speed and direction of the motor in the camera orientation control gimbal 3, achieving precise control of the camera orientation control gimbal 3. An angle sensor (such as a rotary encoder) is installed on the shaft of the motor in the camera orientation control gimbal 3, which can measure the current angular position of the gimbal in real time. When the central processing unit 7-1 sends a motion control command, such as the target angle of the camera orientation control gimbal 3, the integrated controller 7-2 first calculates the difference between the target angle and the current angle. Based on this difference, the integrated controller 7-2 uses a PID (proportional, integral, derivative) control algorithm to adjust the speed and direction of the motor, achieving precise control of the camera orientation. The proportional element directly generates control action based on the magnitude of the difference, causing the motor to quickly approach the target angle; the integral element is used to eliminate the steady-state error of the system, ensuring that the gimbal can accurately stop at the target angle; the derivative element predicts the trend of the system based on the rate of change of the angle, adjusting the control quantity of the motor in advance, making the system more stable and responsive.

[0032] The integrated controller 7-2 controls the camera orientation of the gimbal 3 to achieve fine-tuning of the orientation of each camera 4, further enhancing the flexibility and shooting angle diversity of the entire image acquisition device, thereby meeting the shooting needs in different scenarios. When shooting panoramic images, the integrated controller 7-2 can first control the edge cameras 4 to adjust to a suitable wide-angle shooting angle, and then adjust the angle of the middle cameras 4, ensuring that the shooting areas of each camera 4 are seamlessly connected, thus acquiring a complete, all-around image.

[0033] Optionally, the control system 7 may also include a human-machine interface, through which operators can input control commands and view the working status of the device. The human-machine interface provides operators with a convenient way to operate.

[0034] The working principle and process of this utility model are as follows:

[0035] Initial calibration: First, place the target object to be tested in the detection area. Turn on the laser calibration device and illumination device via the human-machine interface or buttons, and adjust the target object's orientation so that the laser calibration grid completely covers the target object's surface. Simultaneously, each camera acquires an image of the object's surface with the grid, i.e., the grid image, and transmits the grid image to the central processing unit 7-1.

[0036] Adaptive Adjustment: The central processing unit 7-1 calculates the curved laser grid lines in the grid image to achieve calibration conversion from complex curved surfaces to planes. It then analyzes the optimal shooting angle required by each camera, i.e., the target angle, and sends motion control commands to the integrated controller 7-2. The integrated controller 7-2 uses a closed-loop control algorithm (such as PID control algorithm) to calculate the difference between the target angle and the current angle fed back by the angle sensor. Through PID calculation, it adjusts the speed and direction of the motor, driving the camera to rotate precisely toward the control gimbal 3 to the target angle, so that the optical axis of the camera is as perpendicular as possible to the local surface of the target object being photographed. This minimizes image distortion and image shadows caused by the viewing angle until the camera obtains a clear and distinguishable image.

[0037] Image acquisition: After the adjustment is completed, the laser calibration device is turned off, and the central processing unit 7-1 controls all cameras to take pictures simultaneously, acquire clear, non-distorted multi-angle images of the target object's surface, and store the images in the memory 7-3.

[0038] The assembly process of this utility model is as follows:

[0039] First, install support frame 1 and fix its bottom to a horizontal workbench, ensuring that its verticality and horizontality meet the preset requirements;

[0040] Fix the camera, the control gimbal 3, the lighting device 5, and the laser calibration device 6 onto the base plate 2 according to a preset distribution. At the same time, connect the power and signal lines of the camera 4, the power and control lines of the lighting device 5, and the power and control lines of the laser calibration device 6 to the corresponding interfaces of the control system 7.

[0041] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0042] The embodiments described above are merely illustrative of several implementations of this utility model, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the utility model patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this utility model, and these all fall within the protection scope of this utility model. Therefore, the protection scope of this utility model patent should be determined by the appended claims.

Claims

1. A multi-angle image acquisition device for the surface of complex objects based on a multi-camera array, characterized in that, include: Supporting framework (1); A multi-camera imaging system is installed on the support frame (1). The multi-camera imaging system includes a base plate (2), a camera orientation control gimbal (3), and a camera (4) for acquiring images of the surface of a target object. Each camera orientation control gimbal (3) is connected to a camera (4). Multiple camera orientation control gimbals (3) are evenly distributed in an array and rigidly connected to the base plate (2). An illumination device (5) fixed on the base plate (2) is used to provide supplementary light for the camera (4) to acquire images. The illumination device (5) includes at least one elongated light source, which is arranged around the camera array. A laser calibration device (6) fixed on the base plate (2) is used to project a laser calibration grid covering the surface of the target object; The control system (7) includes: The central processing unit (7-1) is communicatively connected to the integrated controller (7-2), the memory (7-3), and each of the cameras (4), respectively, and is used to trigger the cameras (4) to acquire images, receive and process the grid images acquired by each of the cameras (4) when the laser calibration device (6) is turned on, and output motion control commands to the integrated controller (7-2), and receive the images of the surface of the target object acquired by each of the cameras (4) when the laser calibration device (6) is turned off; The integrated controller (7-2) is communicatively connected to the lighting device (5), the laser calibration device (6) and each of the camera orientation control gimbals (3), and is used to control the opening and closing of the lighting device (5) and the laser calibration device (6), and to drive the camera orientation control gimbals (3) to move according to the motion control command; A memory (7-3) communicatively connected to the central processing unit (7-1) is used to store images of the surface of the target object acquired by the camera (4).

2. The multi-camera array-based multi-angle image acquisition device for complex object surfaces according to claim 1, characterized in that, All the camera orientation control gimbals (3) are evenly distributed in a 3×3 array on the base plate (2).

3. The multi-camera array-based multi-angle image acquisition device for complex object surfaces according to claim 1 or 2, characterized in that, The number of elongated light sources is four, and the four elongated light sources are arranged around the camera array.

4. The multi-camera array-based multi-angle image acquisition device for complex object surfaces according to claim 3, characterized in that, The wavelength of the elongated light source is 380-780nm.

5. The multi-camera array-based multi-angle image acquisition device for complex object surfaces according to claim 1 or 2, characterized in that, The camera (4) is a camera with autofocus function.

6. The multi-camera array-based multi-angle image acquisition device for complex object surfaces according to claim 1 or 2, characterized in that, The camera orientation control gimbal (3) is a gimbal with two degrees of freedom: pitch and pan.

7. The multi-camera array-based multi-angle image acquisition device for complex object surfaces according to claim 1 or 2, characterized in that, The laser calibration device (6) uses a structured light generator with a power range of 10-100mW.

8. The multi-camera array-based multi-angle image acquisition device for complex object surfaces according to claim 1 or 2, characterized in that, The laser calibration device (6) emits a laser calibration grid of size 10×10.

9. The multi-camera array-based multi-angle image acquisition device for complex object surfaces according to claim 1 or 2, characterized in that, The integrated controller (7-2) is a programmable logic controller.

10. The multi-camera array-based multi-angle image acquisition device for complex object surfaces according to claim 1 or 2, characterized in that, The camera orientation control gimbal (3) includes a bracket, a motor that drives the bracket to move, and an angle sensor for detecting the rotation angle. The integrated controller (7-2) uses a PID control algorithm to adjust the speed and direction of the motor.

Images