Device and method suitable for demolding and monitoring curved ultra-smooth surface lens

Abstract

The invention discloses an online monitoring device and method for an ultra-smooth surface electroforming demolding interface, and particularly relates to an online monitoring scheme for the demolding process of a high-precision optical element mandrel. The device comprises a supporting mechanism, a conical surface reflecting mirror arranged outside an electroforming mandrel in a sleeving mode, a camera system with an optical axis coinciding with the axis of a rotating body, a dome light source and a processing control unit. The diameter of the small end of the conical reflector is larger than that of the mandrel, so that panoramic non-shielding imaging is ensured. The processing control unit obtains the accurate coordinates of the mark points through a sub-pixel extraction algorithm, and reconstructs a three-dimensional shape and a displacement field based on a ray tracing imaging model and a nonlinear iterative optimization algorithm. The system can be selectively provided with a PSD position sensor and a laser diode, and real-time monitoring of nanoscale changes of a demolding interface can be achieved. According to the invention, 360-degree panorama, submicron precision and real-time three-dimensional tracking monitoring of the separation interface of the mandrel and the element in the electroforming demolding process is realized, and an effective means is provided for precise control of the demolding process of the high-added-value optical element.

Description

Technical Field

[0001] This invention relates to the field of precision optical manufacturing and online inspection technology, specifically to an online monitoring device and method for the demolding interface of ultra-smooth surface electroforming. Background Technology

[0002] In the field of precision optical manufacturing, high-precision aspherical optical components used in high-end laser systems are often formed using electroforming. The final and crucial step in this process is the non-destructive demolding of the cured optical component from the mandrel. Due to the ultra-smooth, high-reflectivity curved surface characteristics of these components, the separation process from the mandrel (i.e., the demolding process) presents challenges such as uneven interface stress distribution, difficulty in capturing the demolding initiation moment, and complex transients during the separation process. Traditional contact sensors or limited visual inspection methods cannot provide 360° panoramic real-time tracking of the demolding interface, making it difficult to avoid component damage or surface shape errors caused by localized stress concentration, resulting in costly component scrapping. Especially for the ultra-smooth curved surfaces, their specular or high-contrast surface characteristics easily cause localized overexposure or shadowing in traditional imaging, severely affecting the consistency and repeatability of observations of key dynamic behaviors such as the demolding initiation point and demolding wave propagation. Therefore, there is an urgent need for an online monitoring solution for the demolding process that can achieve 360° panoramic coverage, uniform illumination, compact structure, and easy calibration.

[0003] Existing single-CCD camera systems have limited fields of view, making it difficult to achieve complete imaging of the mandrel surface demolding process. While multi-CCD systems can expand the detection range, they suffer from problems such as complex synchronization control, high hardware costs, and difficulty in system integration. In addition, traditional pixel-based image processing methods have limited accuracy, making it difficult to meet the precision measurement requirements at the micrometer or even sub-micrometer level. They also lack the ability to highly sense the start of the demolding process, making it impossible to achieve real-time three-dimensional tracking of the demolding interface separation process. Summary of the Invention

[0004] The present invention aims to solve the problems of limited field of view, insufficient accuracy, system complexity and response lag in the existing detection of the demolding process of the curved surface of the rotating body, and provides an optical detection device that can realize 360° panoramic view, high precision and real-time tracking of the demolding interface.

[0005] The technical solution adopted in this invention is:

[0006] A panoramic visual inspection device for rotating bodies based on a conical reflector includes a support mechanism, a conical reflector, an optical imaging system, an illumination system, and a processing and control unit.

[0007] The optical platform is used to fix the rotating sample, ensuring its stable position during the detection process and avoiding mechanical vibration interference;

[0008] The conical reflector is coaxially sleeved outside the electroformed mandrel (9). Its small end diameter is larger than the mandrel diameter, and its vertex is higher than the top of the mandrel. It can fold and converge the reflected light from all directions on the curved surface of the mandrel, thereby realizing 360° panoramic imaging of the entire demolding interface by a single CCD camera.

[0009] The optical imaging system includes a CCD camera and a telecentric lens. The optical axis of the CCD camera coincides with the axis of the mandrel and is positioned towards the conical mirror to acquire panoramic images of the mandrel curved surface converged by the mirror. The telecentric lens plays a crucial role in this system: its unique parallel optical path design effectively eliminates image scaling and distortion caused by changes in object distance, ensuring that images acquired at different heights of the mandrel have a consistent magnification, thereby significantly improving the accuracy of dimensional measurement and 3D reconstruction. In addition, the telecentric lens, used in conjunction with the conical mirror, ensures that each region in the panoramic image is clear and proportionally consistent, providing high-quality input for subsequent image processing and feature extraction.

[0010] The illumination system is located on the outside of the mandrel to provide uniform and stable illumination to the imaging area; preferably, a dome diffuse reflection light source structure is adopted, which can effectively suppress surface reflection and shadows, and improve image quality and feature recognition reliability;

[0011] The processing and control unit is electrically connected to the optical imaging system and is responsible for image acquisition, marker recognition, sub-pixel coordinate extraction, three-dimensional morphology reconstruction, and data analysis. The image processing algorithm used in this device has sub-pixel positioning capability. By extracting and matching image feature points with high precision and combining calibration parameters, it realizes real-time calculation of three-dimensional displacement and strain fields. This algorithm not only improves the spatial resolution of the measurement but also effectively suppresses noise interference and enhances the robustness of the system under complex lighting and surface conditions. Its advantages include the ability to detect micron-level deformation and support dynamic tracking and visualization output of morphology changes throughout the entire demolding and separation process.

[0012] Preferably, the device further includes a laser diode and a position-sensitive detector; the laser diode is used to project a light beam onto a specific monitoring point on the mandrel surface, and the position-sensitive detector receives the reflected light, achieving high-sensitivity detection of nanoscale displacement through optical lever effect; this combination can respond immediately in the initial stage of the demolding process, providing a real-time trigger signal to the processing control unit, ensuring that the system can capture transient changes and perform data correction, and is particularly suitable for monitoring and feedback control of changes at the beginning of the demolding process;

[0013] Furthermore, the device is equipped with an environmental control system, including a vibration isolation platform and a constant temperature control unit, to suppress the impact of external vibrations and temperature fluctuations on detection accuracy and ensure the stability and repeatability of the system in long-term operation.

[0014] The detection method of the present invention includes the following steps:

[0015] S1. The system calibration is completed by using a structured light stereo calibration method based on a phase target. By projecting a multi-frequency phase-shifting grating onto the surface of a precision-machined phase target, a CCD camera is used to acquire the distorted fringe image after being modulated by the curved surface. Combining the principle of stereo vision and the phase-height mapping relationship, a sub-pixel level precision conversion model between pixel coordinates and three-dimensional world coordinates is established.

[0016] S2. Activate the lighting system and CCD camera to acquire panoramic images of the mandrel surface;

[0017] S3. The processing and control unit performs image feature extraction, sub-pixel coordinate calculation, and three-dimensional shape reconstruction in real time using high-precision image processing algorithms based on calibration parameters.

[0018] S4. Dynamic correction is performed by combining feedback data from the laser-position sensor;

[0019] S5. Outputs three-dimensional displacement and strain field information, enabling full-process tracking and precise analysis of the deformation process of the demolding interface.

[0020] Compared with the prior art, the present invention has the following advantages:

[0021] 1. This invention achieves 360° panoramic high-precision imaging of the mandrel curved surface demolding interface by combining a conical reflector and a telecentric lens, effectively solving the problems of limited field of view and error accumulation caused by multiple shots in traditional imaging systems;

[0022] 2. This invention integrates a laser diode and a position-sensitive detector, possessing nanometer-level displacement sensing capabilities, and can capture the starting point of the demolding process in real time, achieving dynamic tracking throughout the entire process;

[0023] 3. This invention employs an image processing algorithm with sub-pixel localization capabilities, combined with uniform illumination and environmental control, significantly improving the accuracy and robustness of 3D shape reconstruction;

[0024] 4. The invention has a compact structure, is easy to calibrate, and is highly adaptable. It can be widely used in demolding monitoring, quality inspection, and scientific research experiments in the manufacturing process of precision optical components, and has good prospects for industrialization. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the image acquisition system of the present invention;

[0026] Figure 2 This is a schematic diagram of the measurement process.

[0027] Figure 3 This is a schematic diagram of the forming relationship of the original part to be tested;

[0028] Figure 4A schematic diagram (bottom view) showing the relative positions of the position-sensitive sensor and the laser diode;

[0029] Figure 5 This is a flowchart of the detection method.

[0030] Explanation of reference numerals: 1. Optical platform; 2. Illumination system lifting mechanism; 3. CCD camera;

[0031] 4. Telecentric lens; 5. Dome diffuse reflection light source; 6. Lighting system support mechanism; 7. Conical reflector; 8. Laser diode; 9. Electroformed mandrel; 10. Processing control unit; 11. Environmental control system; 12. Position-sensitive detector. Detailed Implementation

[0032] To better understand the purpose, structure, and function of this invention, the embodiments of this invention will be described in further detail below with reference to the accompanying drawings.

[0033] like Figures 1-5 As shown, this invention provides an online monitoring device for the demolding process of electroforming ultra-smooth surfaces, used to achieve high-precision, real-time, and all-round monitoring of the dynamic separation behavior of the demolding interface. The device mainly includes an illumination system support mechanism (6), an electroforming mandrel (9), a conical reflector (7), a CCD camera (3), a telecentric lens (4), a dome diffuse reflection light source (5), a processing control unit (10), a laser diode (8), a position-sensitive detector (12), and an environmental control system (11). All components are preferentially mounted on an optical platform (1) to ensure overall rigidity and reduce the impact of vibration.

[0034] The measurement principle of this invention is based on the optical imaging characteristics of a conical reflector, combined with a telecentric lens and uniform illumination, to achieve 360° panoramic imaging and three-dimensional topography reconstruction of the mandrel surface. The optical path is as follows:

[0035] Illumination path: The dome diffuse light source (5) emits uniform diffuse light, illuminating the entire curved surface of the electroformed mandrel (9). The multi-ring LED array on the inner wall of the light source ensures uniform illumination, avoids local overexposure or shadows, and improves image quality.

[0036] Reflected light path: The reflected light from the surface of the mandrel enters the conical reflector (7). The conical reflector is coaxially fitted outside the mandrel, with its small end diameter larger than the mandrel diameter and its cone apex higher than the top of the mandrel. The conical reflector folds and converges the reflected light from all directions of the mandrel sidewalls at a defined angular relationship to form a panoramic light path.

[0037] Imaging optical path: The converged light rays enter the telecentric lens (4). The parallel optical path design of the telecentric lens eliminates image scaling and distortion caused by changes in object distance, ensuring that the image proportions are consistent at different heights. Finally, the light rays form a complete annular panoramic image on the target surface of the CCD camera (3), which contains information about the 360° surface of the core axis.

[0038] 3D Reconstruction Principle: The processing control unit (10) obtains the precise coordinates of the marker points from the image through a sub-pixel extraction algorithm (such as Gaussian surface fitting). Based on the ray tracing imaging model and nonlinear iterative optimization algorithm (such as the Levenberg-Marquardt algorithm), the two-dimensional image coordinates are converted into three-dimensional spatial coordinates to reconstruct the three-dimensional morphology and displacement field of the mandrel surface.

[0039] Example of demolding process monitoring results

[0040] To verify the effectiveness of this invention, we conducted tests in a simulated electroforming demolding experiment. The following are some typical test results:

[0041] Light intensity change detection: In the early stage of the demolding process, image sequences are acquired and light intensity distribution is analyzed. The results show that the system can clearly identify local dark areas or interference fringes caused by air gaps resulting from micro-separation of the mandrel and component interface, with a sensitivity of ±0.1 gray level, achieving accurate determination of the demolding start point.

[0042] Three-dimensional deformation detection: During the demolding process, the reconstructed three-dimensional morphology clearly shows the displacement field distribution of the demolding wavefront propagating along the mandrel axis. The displacement detection accuracy reaches 1μm, and nanometer-level changes are captured in real time by a position-sensitive detector (12) and trigger data correction.

[0043] Real-time tracking results: Throughout the demolding process, the system stably acquires images at a frame rate of 30fps, dynamically outputting displacement vectors and strain tensors. Time series analysis successfully captured the onset and evolution of the demolding wave, with changes in light intensity and deformation field updated synchronously, achieving visualization and quantitative analysis of the entire process.

[0044] These results demonstrate that the present invention can achieve high-precision, panoramic, and real-time detection of the demolding process, and has high sensitivity to physical quantities such as changes in light intensity at the demolding interface.

[0045] The lighting system support mechanism (6) is used to fix and adjust the spatial position and illumination angle of the dome diffuse reflection light source (5) to ensure that the lighting field uniformly covers the entire curved area of ​​the electroformed mandrel (9), while having good mechanical stability to avoid changes in lighting conditions due to displacement or vibration, which would affect the consistency of imaging.

[0046] The support mechanism is also used to fix the electroformed mandrel (9) to ensure that it does not shift or shake during the testing process. The selection of the conical reflector (7) is crucial as a key optical component. The present invention preferably uses a high-precision conical reflector with a surface accuracy better than λ / 4 (@632.8nm) and a surface finish of level III. In practical applications, the cone angle and aperture size of the conical reflector (7) need to be determined according to the diameter and height range of the electroformed mandrel (9) to ensure that it can reflect all the information of the mandrel sidewall to the imaging system without omission. The conical reflector (7) is coaxially fixed to the outside of the electroformed mandrel (9) by a special fixture, which can converge the reflected light from all directions of the curved surface of the electroformed mandrel (9) to the optical imaging system at a certain angle relationship, and finally form a complete, non-overlapping or minimally overlapping annular panoramic image on the target surface of the CCD camera (3), providing a crucial image source for achieving high-precision three-dimensional reconstruction. The conical reflector (7) is fitted outside the electroformed mandrel (9), which can converge the reflected light from all directions of the curved surface of the electroformed mandrel (9) to the optical imaging system to achieve 360° panoramic image acquisition.

[0047] The optical imaging system support mechanism (2) is used to mount and fix the CCD camera (3) and the telecentric lens.

[0048] (4) Ensures precise alignment between the optical axis of the optical imaging system and the axis of the electroformed core shaft (9), and allows for fine-tuning in multiple degrees of freedom to optimize the imaging range and focusing state. The support mechanism has sufficient rigidity and stability to effectively suppress the displacement of the imaging unit caused by external disturbances or its own weight, ensuring spatial consistency and long-term repeatability of image acquisition.

[0049] The optical imaging system consists of a CCD camera (3) and a telecentric lens (4), with the optical axis of the CCD camera (3) strictly aligned with the axis of the electroformed core shaft (9). The use of the telecentric lens (4) effectively overcomes the image distortion and size distortion caused by inconsistent object distances in traditional lenses, ensuring uniform proportions throughout the panoramic imaging and providing an accurate and consistent image basis for subsequent 3D reconstruction.

[0050] The lighting system uses a dome diffuse reflection light source (5), which reduces surface reflection and shadows through uniform illumination, significantly improving image quality and feature recognition stability.

[0051] The processing control unit (10) receives image data acquired by the CCD camera (3) and uses a high-precision image processing algorithm to achieve marker point recognition, sub-pixel coordinate extraction, and three-dimensional shape reconstruction. This algorithm has good noise suppression capability and sub-pixel positioning accuracy, and can generate displacement field and strain field data of the demolding interface on the surface of the electroformed mandrel (9) in real time, and dynamically track the entire demolding separation process, providing a reliable basis for achieving precise control and optimization of the demolding process.

[0052] The laser diode (8) and the position-sensitive detector (12) work together to form a high-sensitivity displacement sensing unit. The laser diode (8) projects a light spot onto the surface of the electroformed mandrel (9), and the position-sensitive detector (12) receives the reflected light signal, detecting nanoscale displacement by the change in the position of the light spot. This unit can capture the start moment of the demolding process in a timely manner and feed back a signal to the processing control unit (10) to trigger real-time data correction and process control, making it suitable for high-precision machining scenarios that are sensitive to the demolding process.

[0053] The environmental control system (11) includes a vibration isolation platform and a constant temperature unit, which effectively isolates external vibrations and maintains a stable temperature environment, ensuring that the system maintains high repeatability and measurement accuracy during long-term operation.

[0054] The system workflow is as follows:

[0055] S1. The system calibration was completed by using a structured light stereo calibration method based on a phase target. A four-step phase-shift sinusoidal grating and a Gray code pattern were sequentially projected onto the surface of a precision-machined phase target. A CCD camera (3) was used to collect the distorted stripe image after being modulated by the curved surface. The phase unwrapping was performed using the multi-frequency heterodyne principle. Combining the stereo vision principle with the phase-height mapping relationship, a sub-pixel level precision conversion model between pixel coordinates and three-dimensional world coordinates was established by least squares fitting. The calibration accuracy reached 0.05 pixels.

[0056] S2. Start the dome diffuse reflection light source (5) to provide uniform illumination, and simultaneously trigger the CCD camera (3) to acquire the panoramic image reflected by the conical reflector (7) on the surface of the electroformed mandrel (9); set the image acquisition frame rate to 30fps and the resolution to 2048×2048 pixels to ensure that the dynamic change process of the demolding interface can be captured.

[0057] S3. The processing control unit (10) performs image preprocessing based on calibration parameters using a high-precision image processing algorithm, eliminates noise interference by using Gaussian filtering, and extracts edge features using the Canny operator; it obtains feature point coordinates by sub-pixel positioning technology based on gray-scale centroid method, and calculates surface morphology by combining stereo vision three-dimensional reconstruction algorithm, achieving a reconstruction accuracy of 1μm;

[0058] S4. Synchronously acquire the voltage signal output by the position sensitive detector (12), and calculate the laser spot offset in real time through the four-quadrant position detection principle; when a nanometer-level change is detected at the demolding interface, immediately send a trigger signal to the processing control unit (10), and dynamically correct the three-dimensional reconstruction result based on the Kalman filter algorithm to eliminate the measurement error caused by environmental disturbance;

[0059] S5. Outputs complete deformation field data including three-dimensional coordinates, displacement vectors, and strain tensors, displaying the deformation distribution of the demolding interface in point cloud format and color cloud map form; simultaneously, it outputs light intensity change images and displacement-time curves, enabling full-process tracking and precise analysis of the dynamic separation process of the demolding interface. The measurement capabilities of the system are intuitively demonstrated through the detection results (such as light intensity distribution maps, three-dimensional morphology reconstruction maps, and displacement-time curves).

[0060] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A device and method for online monitoring of the demolding interface in ultra-smooth surface electroforming, characterized in that, It includes a support mechanism (6), a conical reflector (7), an optical imaging system, an illumination system, and a processing control unit (10); the conical reflector (7) is coaxially mounted outside the electroformed mandrel (9); the optical axis of the optical imaging system coincides with the axis of the electroformed mandrel (9) and is set toward the conical reflector (7); the processing control unit (10) is electrically connected to the optical imaging system and is used to process images and reconstruct the three-dimensional morphology of the demolding interface.

2. The apparatus according to claim 1, characterized in that: The inner diameter of the small end of the conical reflector (7) is greater than the maximum outer diameter of the electroformed mandrel (9), and the cone apex of the conical reflector (7) is higher than the top of the electroformed mandrel (9).

3. The apparatus according to claim 1, characterized in that: The optical imaging system includes a CCD camera (3) and a telecentric lens (4).

4. The apparatus according to claim 1, characterized in that: The lighting system is a dome diffuse light source (5), which includes a hemispherical diffuser and a multi-ring LED array disposed on its inner wall to provide uniform diffuse lighting.

5. The apparatus according to claim 1, characterized in that: The processing control unit (10) is configured to: extract the sub-pixel image coordinates of the marker points from the image acquired by the optical imaging system using a sub-pixel extraction algorithm; convert the sub-pixel image coordinates into three-dimensional spatial coordinates using a nonlinear iterative optimization algorithm based on the ray tracing imaging model; and reconstruct the three-dimensional morphology and displacement field of the surface of the electroformed mandrel (9) according to the time series of the three-dimensional spatial coordinates.

6. The apparatus according to claim 5, characterized in that: The subpixel extraction algorithm includes at least one of Gaussian surface fitting and gray-level squared weighted centering method; the nonlinear iterative optimization algorithm adopts the Levenberg-Marquardt algorithm.

7. The apparatus according to claim 1, characterized in that: It also includes a position-sensitive detector (12) and a laser diode (8); the laser diode (8) is used to project a probe beam onto the surface of the electroformed mandrel (9); the position-sensitive detector (12) is used to receive the reflected beam to detect nanoscale displacement.

8. The apparatus according to claim 7, characterized in that: The processing control unit (10) is further configured to generate a trigger signal based on the rate of change of the output signal of the position-sensitive detector (12).

9. A method of using the apparatus according to any one of claims 1 to 8, characterized in that, Includes the following steps: S1, System calibration: The system calibration was completed using a structured light stereo calibration method based on a phase target; S2, Synchronous Acquisition: Synchronously trigger the lighting system and optical imaging system to acquire a panoramic image sequence of the surface of the electroformed mandrel (9); S3, Subpixel Extraction and 3D Reconstruction: The processing control unit (10) extracts the subpixel coordinates of the marker points and reconstructs their 3D coordinates based on the calibration model; S4, Deformation Analysis: Based on the time series of three-dimensional coordinates, calculate the three-dimensional displacement field and deformation field of the surface of the electroformed mandrel (9); S5, Output Results: Displays and outputs the dynamic three-dimensional morphology and quantitative data of the demolding process. A non-transitory computer-readable storage medium storing a computer program, which, when executed by a processing control unit (10), performs the steps of the method as described in claim 9.

Images