A system and method for three-dimensional reconstruction of object shape and deformation using laser speckle

Abstract

The application discloses a system and method for three-dimensional reconstruction of object shape and deformation by using laser speckle, and belongs to the field of optical measurement and three-dimensional imaging. The system forms a random and uniform speckle field on the surface of an object through a speckle laser emission module, synchronously collects speckle images by using a multi-view high-speed camera, obtains accurate parameters after camera calibration, tracks and matches speckle feature points by using a three-dimensional PTV algorithm, and solves three-dimensional coordinates and dynamic displacement vectors of the object. The technology does not need to manually mark feature points, realizes contactless and integrated measurement, has high precision and high time resolution at a sub-pixel level, can capture nanoscale tiny deformation and complex curved surface dynamic processes, has strong anti-interference ability, is suitable for industrial detection, structure health monitoring, experimental mechanics measurement and other scenes, effectively solves the problems of separation of static and dynamic measurement, complicated operation and narrow application range of traditional technologies, and improves measurement efficiency and data reliability.

Description

Technical Field

[0001] This invention relates to the field of optical measurement and three-dimensional imaging technology, specifically to a system and method for three-dimensional reconstruction of the shape and deformation of an object using laser speckle. Background Technology

[0002] Traditional 3D morphology deformation measurement techniques, such as optical imaging approaches like structured light 3D scanning, can achieve high-precision reconstruction of static surface morphology. However, they often have limitations when facing dynamic deformation measurement and detection of minute deformations, and are extremely sensitive to environmental vibrations, with a limited measurement range. Another approach involves adding feature points distributed across the model surface for identification. The 3D coordinates of these feature points are first reconstructed, and then the morphology and displacement of the model surface are mapped. This method requires surface spraying / etching markings, which is extremely difficult to implement in many scenarios, such as assembly line workpieces, large targets (e.g., bridges, buildings), and extreme conditions (e.g., underwater, high temperatures). Furthermore, the feature point markings themselves may affect the measurement results.

[0003] The core technical challenge of laser speckle-based 3D measurement methods lies in how to stably and accurately extract the 3D shape information of an object and its changes over time from random, high-noise speckle images. Existing speckle correlation or phase analysis methods often face challenges in achieving full-field, real-time, and high-precision 3D dynamic deformation measurement, including high computational complexity, weak anti-interference capabilities, difficulties in multi-view synchronization, and high requirements for the optical properties of the object's surface. Furthermore, existing systems still suffer from bottlenecks in achieving miniaturization, integration, and multi-parameter synchronous measurement, such as complex hardware layout, cumbersome adjustments, and insufficient environmental adaptability.

[0004] To address the above issues, especially to achieve rapid, stable, and high-precision 3D reconstruction of the shape and deformation of objects under complex curved surfaces, dynamic processes, or small scales, it is necessary to develop a novel laser speckle measurement system architecture and processing algorithm. Summary of the Invention

[0005] The purpose of this invention is to provide a system and method for three-dimensional reconstruction of the shape and deformation of an object using laser speckle, in order to solve the problems mentioned in the background art, such as the separation of static shape and dynamic deformation measurement, the need for manual marking of feature points, weak anti-interference ability, and difficulty in achieving high-precision real-time measurement of complex curved surfaces and small deformations.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a system for three-dimensional reconstruction of the shape and deformation of an object using laser speckle, comprising:

[0007] Speckle laser emission module: Composed of a speckle laser emitter and a fixture, it is used to apply a uniformly covered speckle laser spot to the surface of the object to be tested, and the speckle laser itself serves as the illumination source;

[0008] Multi-view high-synchronization image acquisition module: It consists of at least two high-speed cameras fixed on a rigid bracket and a camera calibration program, used to synchronously acquire speckle images of the surface of the object under test and obtain camera parameters;

[0009] Data processing and 3D reconstruction solution module: used to identify, match and correct feature points in the acquired multi-view speckle images, track speckle patterns across time series using the 3D PTV algorithm, solve the 3D displacement vector of each point on the object surface, and complete the 3D morphology and dynamic deformation reconstruction.

[0010] Preferably, the speckle laser emitter is fixed to the optical platform by a clamp, and its position, angle and focal length can be adjusted so that the speckle laser points on the object surface reach the minimum and brightest state.

[0011] Preferably, the baseline distance of the high-speed camera does not exceed 20cm, and the camera calibration program calculates the internal and external parameters of the camera by taking 10 to 20 sets of images of the asymmetric dot calibration plate in different poses.

[0012] This invention also provides a method for three-dimensional reconstruction of the shape and deformation of an object using laser speckle. This method is based on the above-mentioned system, and the specific steps are as follows:

[0013] Step 1: System installation and alignment. Fix and debug the speckle laser emission module and the multi-view high-synchronization image acquisition module to ensure that the laser uniformly covers the area to be measured and the camera's field of view completely covers the object to be measured.

[0014] Step 2: Camera calibration. Obtain and save the camera parameters using a calibration board for subsequent 3D calculations.

[0015] Step 3: Speckle image acquisition. Remove the calibration plate and simultaneously capture a speckle image of the surface of the object under test using a high-speed camera.

[0016] Step 4: Data processing and reconstruction. The speckle image is preprocessed by grayscale threshold extraction. The speckle pattern is tracked and matched using the 3D PTV algorithm. The 3D displacement vector is calculated to reconstruct the 3D shape and dynamic deformation data of the object, and then further analysis is performed.

[0017] Preferably, the three-dimensional PTV algorithm described in step 4 is used to track speckle patterns modulated by the microstructure of the object surface, forming a spatiotemporally unified four-dimensional deformable data volume, and realizing sub-pixel or even nanometer-level motion measurement.

[0018] Preferably, in step 4, the displacement-time curve of a specific point can be analyzed, the strain distribution of the entire field can be calculated, and the evolution of the deformation mode can be observed based on the reconstructed speckle spatial coordinates.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0020] 1. Breakthrough in both measurement accuracy and dynamic performance. By leveraging the 3D PTV algorithm to accurately track speckle feature points, sub-pixel-level 3D coordinate calculation and micro-deformation recognition are achieved. At the same time, dynamic data acquisition is supported, which solves the bottleneck of traditional technologies that cannot simultaneously achieve "high-precision static shape reconstruction" and "high temporal resolution dynamic deformation tracking". It can completely capture the entire process of dynamic deformation of objects from micro to macro.

[0021] 2. Significantly improved operational efficiency. No need to manually design or add feature points or physical tracers on the object's surface; laser speckle can quickly and uniformly cover the surface to be measured. Combined with automated calibration and image acquisition processes, it integrates the traditional two-step operation of "static scanning + dynamic measurement," shortening the overall measurement time and eliminating the need for tedious adjustments by professional personnel, thus lowering the operational threshold.

[0022] 3. The system has a simple hardware layout, requiring no additional complex optical components, which reduces equipment costs; the measurement process is non-contact and non-destructive, and will not change the material properties of the object surface or interfere with the measurement process, making it widely applicable in fields such as industrial inspection, structural health monitoring, and experimental mechanics. Attached Figure Description

[0023] Figure 1 The diagrams show the speckle pattern processing; (a) the original speckle pattern from the left camera; (b) the original speckle pattern from the right camera; (c) the preprocessed speckle pattern from the left camera; and (d) the preprocessed speckle pattern from the right camera.

[0024] Figure 2 This is a schematic diagram of the 3D reconstruction result. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Implementation preparation:

[0027] Hardware preparation: Assemble a measurement system consisting of a laser emission module, a multi-view image acquisition module (including at least two high-speed cameras fixed on a rigid bracket, and a calibration board for camera calibration, preferably an asymmetric dot calibration board).

[0028] Software preparation: Install a program that integrates a camera and 3D reconstruction on the computer.

[0029] The software must include the following core functional modules:

[0030] Camera control module: Supports simultaneous triggering of two cameras and allows setting shooting parameters;

[0031] Calibration calculation module: Built-in calibration algorithm, supports automatic identification of calibration board dots, calculation of camera intrinsic parameters (focal length, principal point, distortion coefficient) and extrinsic parameters (relative position, attitude angle);

[0032] Image processing module: integrates grayscale threshold extraction, speckle noise reduction, and feature point matching algorithms;

[0033] 3D Reconstruction and Analysis Module: Equipped with a 3D PTV algorithm, it can calculate 3D coordinates and displacement vectors, and supports the generation of displacement-time curves and full-field strain contour maps.

[0034] Implementation steps:

[0035] Step 1: System Installation and Initial Alignment

[0036] Fix the speckle laser probe to a suitable position on the optical platform using a clamp, ensuring it faces the area to be measured. Adjust the position and angle of the probe to ensure that the field of view of all cameras completely covers the object under test, and that the laser illumination is uniformly distributed over the area. Adjust the laser focal length to minimize and maximize the speckle laser spots on the object's surface.

[0037] Step 2: Camera Installation and Calibration

[0038] Install two cameras that meet the experimental requirements in suitable locations, with the camera baseline within 20cm for easy matching. Place a high-precision 3D calibration board (such as an asymmetric dot calibration board) within the spatial range of the object under test, and vary its orientation and angle multiple times. Simultaneously capture multi-view images of the calibration board in different orientations using all cameras (typically 10-20 sets of images are required). The software uses the captured calibration images to automatically calculate and optimize the internal parameters (focal length, principal point, distortion coefficient) of each camera and the external parameters (relative position and orientation) between the cameras. After calibration, the software saves the calibration parameter file, which will be used for all subsequent 3D calculations.

[0039] Step 3: Acquire speckle laser images

[0040] Remove the calibration plate and place the object to be measured at the measurement position. Simultaneously capture images of the model surface using all cameras, ensuring that the resulting speckle images are clear and not overexposed.

[0041] Step 4: Post-measurement processing and analysis

[0042] The original speckle pattern is subjected to grayscale thresholding to extract laser points. A reconstruction program is then used to perform 3D reconstruction of the preprocessed image. The saved spatial coordinates of the speckles allow for analysis of displacement-time curves at specific points, calculation of overall strain distribution, and observation of deformation mode evolution.

[0043] Figure 1 is a schematic diagram of speckle pattern processing, containing four sub-images: (a) original speckle pattern from the left camera, (b) original speckle pattern from the right camera, (c) preprocessed speckle pattern from the left camera, and (d) preprocessed speckle pattern from the right camera, visually presenting the changes in the image from its original state to its processed state that can be used for calculation.

[0044] Preprocessing is an initial optimization operation performed on the acquired raw speckle image. The core of it is grayscale threshold extraction and speckle noise reduction. The purpose is to filter image noise and highlight speckle feature points, so that the subsequent 3D PTV algorithm can more accurately track and match the speckle pattern, laying the foundation for 3D coordinate calculation and deformation reconstruction.

[0045] Figure 2 is a schematic diagram of the three-dimensional reconstruction results, which mainly presents the three-dimensional spatial morphology and dynamic deformation data of the object under test after measurement using the laser speckle system.

[0046] Based on the key information in the figure, it marks the spatial dimensions by coordinate axes (Z-axis, etc.) and combines numerical matrices to quantify the three-dimensional coordinates or displacement of each point on the object's surface, intuitively reflecting the object's three-dimensional contour, surface undulations and deformation distribution, which can assist in subsequent in-depth processing such as specific point displacement analysis and full-field strain calculation.

[0047] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A system for three-dimensional reconstruction of the shape and deformation of an object using laser speckle, characterized in that, include: Speckle laser emission module: It consists of a diffractive speckle laser emitter and a fixture, and is used to apply a uniformly covered speckle laser spot to the surface of the object to be tested, with the speckle laser itself serving as the illumination source; Multi-view high-synchronization image acquisition module: It consists of at least two high-speed cameras fixed on a rigid bracket and a camera calibration program, used to synchronously acquire speckle images of the surface of the object under test and obtain camera parameters; Data processing and 3D reconstruction solution module: used to identify, match and correct feature points in the acquired multi-view speckle images, track speckle patterns across time series using the 3D PTV algorithm, solve the 3D displacement vector of each point on the object surface, and complete the 3D morphology and dynamic deformation reconstruction.

2. The system according to claim 1, characterized in that, The speckle laser emitter is fixed to the optical platform by a clamp, and its position, angle and focal length can be adjusted to make the speckle laser points on the object surface reach the smallest and brightest state.

3. The system according to claim 1, characterized in that, The camera calibration program calculates the camera's internal and external parameters by taking 10 to 20 sets of calibration board images in different poses.

4. A method for three-dimensional reconstruction of the shape and deformation of an object using laser speckle, characterized in that, The specific steps of using the system as described in any one of claims 1-3 are as follows: Step 1: System installation and alignment. Fix and debug the speckle laser emission module and the multi-view high-synchronization image acquisition module to ensure that the laser uniformly covers the area to be measured and the camera's field of view completely covers the object to be measured. Step 2: Camera calibration. Obtain and save the camera parameters using a calibration board for subsequent 3D calculations. Step 3: Speckle image acquisition. Remove the calibration plate and simultaneously capture a speckle image of the surface of the object under test using a high-speed camera. Step 4: Data processing and reconstruction. The speckle image is preprocessed by grayscale threshold extraction. The speckle pattern is tracked and matched using the 3D PTV algorithm. The 3D displacement vector is calculated to reconstruct the 3D shape and dynamic deformation data of the object, and then further analysis is performed.

5. The method according to claim 4, characterized in that, The three-dimensional PTV algorithm described in step 4 is used to track speckle patterns modulated by the microstructure of the object's surface, forming a spatiotemporally unified four-dimensional deformable data volume to achieve sub-pixel motion measurement.

6. The method according to claim 4, characterized in that, In step 4, based on the reconstructed speckle spatial coordinates, the displacement-time curves of specific points can be analyzed, the strain distribution across the entire field can be calculated, and the evolution of deformation modes can be observed.

Images