Non-contact Measurement Technique Research Articles

Algorithm Analysis and Optimization of a Digital Image Correlation Method Using a Non-Probability Interval Multidimensional Parallelepiped Model

Digital image correlation (DIC), a widely used non-contact measurement technique, often requires empirical tuning of several algorithmic parameters to strike a balance between computational accuracy and efficiency. This paper introduces a novel uncertainty analysis approach aimed at optimizing the parameter intervals of a DIC algorithm. Specifically, the method leverages the inverse compositional Gauss–Newton algorithm combined with a prediction-correction scheme (IC-GN-PC), considering three critical parameters as interval variables. Uncertainty analysis is conducted using a non-probabilistic interval-based multidimensional parallelepiped model, where accuracy and efficiency serve as the reliability indexes. To achieve both high computational accuracy and efficiency, these two reliability indexes are simultaneously improved by optimizing the chosen parameter intervals. The optimized algorithm parameters are subsequently tested and validated through two case studies. The proposed method can be generalized to enhance multiple aspects of an algorithm’s performance by optimizing the relevant parameter intervals.

Penetrating Barriers: Noncontact Measurement of Vital Bio Signs Using Radio Frequency Technology.

The noninvasive measurement and sensing of vital bio signs, such as respiration and cardiopulmonary parameters, has become an essential part of the evaluation of a patient's physiological condition. The demand for new technologies that facilitate remote and noninvasive techniques for such measurements continues to grow. While previous research has made strides in the continuous monitoring of vital bio signs using lasers, this paper introduces a novel technique for remote noncontact measurements based on radio frequencies. Unlike laser-based methods, this innovative approach offers the advantage of penetrating through walls and tissues, enabling the measurement of respiration and heart rate. Our method, diverging from traditional radar systems, introduces a unique sensing concept that enables the detection of micro-movements in all directions, including those parallel to the antenna surface. The main goal of this work is to present a novel, simple, and cost-effective measurement tool capable of indicating changes in a subject's condition. By leveraging the unique properties of radio frequencies, this technique allows for the noninvasive monitoring of vital bio signs without the need for physical contact or invasive procedures. Moreover, the ability to penetrate barriers such as walls and tissues opens new possibilities for remote monitoring in various settings, including home healthcare, hospital environments, and even search and rescue operations. In order to validate the effectiveness of this technique, a series of experiments were conducted using a prototype device. The results demonstrated the feasibility of accurately measuring respiration patterns and heart rate remotely, showcasing the potential for real-time monitoring of a patient's physiological parameters. Furthermore, the simplicity and low-cost nature of the proposed measurement tool make it accessible to a wide range of users, including healthcare professionals, caregivers, and individuals seeking to monitor their own health.

Statistical blade tip timing measurement, Part II: High-order cumulant architecture

Blade tip timing (BTT) is a potential non-contact vibration measurement technique for rotating blades owing to its high efficiency and long service life. However, probe layout design and vibration parameter identification are critical challenges in BTT due to its inherent undersampling. To address this issue, we have proposed the concept of statistical BTT measurement which promotes efficient layouts and parameter identification methods by shifting the recovery target from the signal to itself to its statistics. In the first paper of series papers, we have developed covariance architecture for BTT measurement and signal post-processing. In this second part of series papers, we set out to develop BTT measurement from a higher-order statistic perspective. Specifically, we find that the 2qth-order cumulant retains the vibration information (frequency and amplitude) of the signal itself and it can be recovered at lower sampling rates than what is prescribed by the covariance architecture. On this basis, we propose a higher-order cumulant architecture for BTT measurement, which contains two aspects: higher-order difference layout (HODL) and higher-order cumulant-based parameter identifications. HODL is a special family of layouts that physically guarantee the availability of consecutive higher-order cumulant estimations and thus leads to a satisfactory parameter identification. To facilitate application, a series of concrete HODLs are provided for users, including a four-level nested layout and its enhanced version, and optimal fourth-order difference layout. The quantitative comparison in the simulations and experiments shows the effectiveness of the proposed higher-order cumulant architecture. Compared with the architectures developed from covariance and signal itself perspective, the proposed higher-order cumulant architecture is not only more robust to noise but also further reduces the number of required probes.

Anthropometric Study of Correlation between the Selected Craniofacial Measurements with the Stature in Adult Bangladeshi Manipuri Males

Cephalofacial anthropometry can be useful for racial identification, forensic scientist, physical anthropologist, genetic counsellors as well as reconstructive surgery for the purposes of identification of an individuals and understanding human physical variation, gender and ethnicity especially with facial recognition as a tool in recent advances in biometrics. Stature is one of the important criteria for personal identification which has a proportional biological relationship with every parts of the human body like head, face, trunk, extremities and vertebral column. In this study, our main goal is find out the correlation between the selected craniofacial measurements with the stature in Adult Bangladeshi Manipuri males. This cross-sectional study was carried out in 100 healthy adult males from March 2017 to February 2018 in the Department of Anatomy, Bangabandhu Sheikh Mujib Medical University (BSMMU), Dhaka, Bangladesh. Descriptive statistics and testing of hypotheses were used for the analysis using SPSS software (version 22.0). Here non- significant positive correlations of the stature with the mandible height and significant positive correlations of the stature with mandible breadth was found. The Mean (±SD) of Mandible height (sto-gn) was 4.44 (±0.69), Mandible breadth (go-go) was 11.45 (±0.57). This study will be anticipated to provide baseline quantitative data on the linear craniofacial measurements and the stature of adult Bangladeshi Manipuri males. And using larger samples with non-contact measurement technique will help in defining craniofacial anthropometric profiling of the adult Bangladeshi Manipuri males. Mugda Med Coll J. 2024; 7(1): 19-25

Integrated Neural Network Approach for Enhanced Vital Signal Analysis Using CW Radar

This study introduces a novel approach for analyzing vital signals using continuous-wave (CW) radar, employing an integrated neural network model to overcome the limitations associated with traditional step-by-step signal processing methods. Conventional methods for vital signal monitoring, such as electrocardiograms (ECGs) and sphygmomanometers, require direct contact and impose constraints on specific scenarios. Conversely, our study primarily focused on non-contact measurement techniques, particularly those using CW radar, which is known for its simplicity but faces challenges such as noise interference and complex signal processing. To address these issues, we propose a temporal convolutional network (TCN)-based framework that seamlessly integrates noise removal, demodulation, and fast Fourier transform (FFT) processes into a single neural network. This integration minimizes cumulative errors and processing time, which are common drawbacks of conventional methods. The TCN was trained using a dataset comprising preprocessed in-phase and quadrature (I/Q) signals from the CW radar and corresponding heart rates measured via ECG. The performance of the proposed method was evaluated based on the L1 loss and accuracy against the moving average of the estimated heart rates. The results indicate that the proposed approach has the potential for efficient and accurate non-contact vital signal analysis, opening new avenues in health monitoring and medical research. Additionally, the integration of CW radar and neural networks in our framework offers a robust and scalable solution, enhancing the practicality of non-contact health monitoring systems in diverse environments. This technology can be leveraged in healthcare robots to provide continuous and unobtrusive monitoring of patients’ vital signs, enabling timely interventions and improving overall patient care.

Energy Efficiency Measurement of Mechanical Crushing Based on Non-Contact Identification Method

The efficiency of mechanical crushing is a key metric for evaluating machinery performance. However, traditional contact-based methods for measuring this efficiency are unable to provide real-time data monitoring and can potentially disrupt the production process. In this paper, we introduce a non-contact measurement technique for mechanical crushing efficiency based on deep learning algorithms. This technique utilizes close-range imaging equipment to capture images of crushed particles and employs deeply trained algorithmic programs rooted in symmetrical logical structures to extract statistical data on particle size. Additionally, we establish a relationship between particle size and crushing energy through experimental analysis, enabling the calculation of crushing efficiency data. Taking cement crushing equipment as an example, we apply this non-contact measurement technique to inspect cement particles of different sizes. Using deep learning algorithms, we automatically categorize and summarize the particle size ranges of cement particles. The results demonstrate that the crushing efficiencies of ore crushing particles, raw material crushing particles, and cement crushing particles can respectively reach 80.7%, 70.15%, and 80.27%, which exhibit a high degree of consistency with the rated value of the samples. The method proposed in this paper holds significant importance for energy efficiency monitoring in industries that require mechanical crushing.

A Novel Simulation Method for 3D Digital-Image Correlation: Combining Virtual Stereo Vision and Image Super-Resolution Reconstruction.

3D digital-image correlation (3D-DIC) is a non-contact optical technique for full-field shape, displacement, and deformation measurement. Given the high experimental hardware costs associated with 3D-DIC, the development of high-fidelity 3D-DIC simulations holds significant value. However, existing research on 3D-DIC simulation was mainly carried out through the generation of random speckle images. This study innovatively proposes a complete 3D-DIC simulation method involving optical simulation and mechanical simulation and integrating 3D-DIC, virtual stereo vision, and image super-resolution reconstruction technology. Virtual stereo vision can reduce hardware costs and eliminate camera-synchronization errors. Image super-resolution reconstruction can compensate for the decrease in precision caused by image-resolution loss. An array of software tools such as ANSYS SPEOS 2024R1, ZEMAX 2024R1, MECHANICAL 2024R1, and MULTIDIC v1.1.0 are used to implement this simulation. Measurement systems based on stereo vision and virtual stereo vision were built and tested for use in 3D-DIC. The results of the simulation experiment show that when the synchronization error of the basic stereo-vision system (BSS) is within 10-3 time steps, the reconstruction error is within 0.005 mm and the accuracy of the virtual stereo-vision system is between the BSS's synchronization error of 10-7 and 10-6 time steps. In addition, after image super-resolution reconstruction technology is applied, the reconstruction error will be reduced to within 0.002 mm. The simulation method proposed in this study can provide a novel research path for existing researchers in the field while also offering the opportunity for researchers without access to costly hardware to participate in related research.

Promising lanthanide- NaYbF4:Er3+@NaYbF4:Tm3+ micro-phosphors for highly efficient upconversion luminescence and temperature sensing

In recent years, the fluorescence intensity ratio (FIR)-based technique for optical non-contact temperature measurement is of great interest to researchers due to its promising wide range of applications in high-voltage electricity fields, fires, corrosive environments, biological tissues, as well as other fields. Unfortunately, the FIR techniques based on thermally coupled energy levels (TCLs) often have the problem of low sensitivity as a result of the narrow energy gap. In contrast to FIR techniques with TCLs, the FIR technique with non-thermal coupled energy levels (NTCLs) is no longer restricted by the energy gap and is expected to achieve high signal resolution. In this study, we attempted to acquire a thermometric material that possesses high sensitivity as well as high signal resolution with the help of NTCLs. We successfully prepared NaYbF4:Er3+@NaYbF4:Tm3+ core-shell microcrystals by solvothermal technique using NaYbF4 as matrix and Er3+ and Tm3+ as luminescent centers and made necessary investigations on their crystal structures, microscopic morphologies, intrinsic mechanisms, and thermometric properties. The sample phosphors in a 980 nm laser show bright upconversion (UC) luminescence with emission peaks corresponding to the characteristic energy level jumps of Er3+ as well as Tm3+, respectively. The temperature sensing characteristic parameters have been investigated by FIR techniques with TCLs (2H11/2 → 4I15/2(Er3+)/4S3/2 → 4I15/2(Er3+)) & NTCLs (3H4→3H6(Tm3+)/1G4→3H6(Tm3+),3H4→3H6(Tm3+)/2H11/2 → 4I15/2(Er3+),3H4→3H6(Tm3+)/4S3/2 → 4I15/2(Er3+) and 3H4→3H6(Tm3+)/4F9/2 → 4I15/2(Er3+)), respectively. It is shown that the maximum absolute & relative sensitivities are 0.5677 K-1(495 K) as well as 1.20 % K−1(445 K) respectively with good thermal stability. In addition, the NTCLs have good separation of the emission bands, which enables the material to have excellent signal discrimination in temperature detection. These results indicate that the material is expected to be a potential candidate for optical thermometers.

Research on the application of DIC technology in model tests of hydraulic structures

Digital image correlation (DIC) is a noncontact measurement technique that can visualize and measure a model body in an entire field without interference. This study performed the structural model test of a cemented sand and gravel (CSG) dam. The scale-free DIC technology was applied to the model test to monitor the CSG model dam during overload damage. The study results demonstrated the following results:1) The CSG dam experienced sliding shear failure along the dam foundation with an overload coefficient of 5.2 during failure. 2) The use of DIC technology for measurement could reflect the generation and development of the damage process of the model dam. 3) The DIC technique used in this test achieved an observation accuracy of 2 mm for displacement and 530 με for strain. 4) The conclusions of the measurement results of the DIC technique and the traditional method are identical. The DIC technique has considerable application potential in the model tests of hydro-structures. This study provides a new measurement method for hydraulic structure model tests and supports the development of an intelligent monitoring platform for subsequent hydraulic structure model tests.

A recursive calculation method of vibration displacements using blade tip timing in angular domain

Blade tip timing (BTT) is a non-contact measurement technique that can be used to monitor blade vibration in turbomachinery. The raw data measured by BTT is the arrival time of all blades at the same stage. However, most blade vibration monitoring methods based on BTT require the use of blade vibration displacements, which are derived from the raw arrival time data, the blade rotation radius and speed. Conventional BTT methods calculate blade tip displacements by comparing the measured arrival time with the ideal arrival time. The ideal arrival time is obtained by assuming that the rotational speed is constant in a revolution, which is inconsistent with the fact, especially when the rotational speed changes rapidly. In this paper, a recursive calculation method called recursive BTT is proposed to reduce the influence of rotational speed fluctuation on the accuracy of blade tip displacement calculation. Recursive BTT calculates the displacements in the angular domain instead of the time domain. A recursive formula is constructed in the angular domain to calculate the blade vibration displacement by using the angle between the vibrating blades. Compared with the conventional BTT method, which relies on the once-per-revolution (OPR) signal, recursive BTT makes full use of the arrival time of all blades at the same stage to calculate the displacements. The advantage of recursive BTT is illustrated through a simulation, a laboratory test and a full-scale aeroengine test. To evaluate the reliability of the calculated displacements, strain signals are resampled to compare with the calculated displacements in the laboratory test. In the full-scale aeroengine test, due to the lack of strain signals, BTT displacements are used to track the blade natural frequency to qualitatively indicate the influence of displacements on blade vibration monitoring. The results of finite element analysis and modal test are used for verification in the full-scale aeroengine test. Simulation and experimental results demonstrate that compared with existing methods, recursive BTT is robust for calculating blade vibration displacements.

Non-Contact Evaluation of Deformation Characteristics on Automotive Steel Sheets

Non-Contact Evaluation of Deformation Characteristics on Automotive Steel Sheets

Stereo-DICNet: An efficient and unified speckle matching network for stereo digital image correlation measurement

Digital image correlation (DIC) is an image-based, non-contact optical measurement technique for measuring the deformation of objects. Recently, deep learning has been preliminarily used in monocular-DIC measurement and has brought significant advantages in efficiency and robustness. However, the deep learning-based stereo-DIC methods is still underexplored, although stereo-DIC can measure three-dimensional deformation and shape compared to monocular-DIC. Different from monocular-DIC, stereo-DIC not only requires temporal matching (inter-frame) but also stereo matching (inter-camera), and the difference in image deformation scale between the two matching tasks make it difficult to perform them using a unified network. To solve this problem, we propose a unified speckle matching network that can be used for deformation measurement at different scales, referred to as Stereo-DICNet. Stereo-DICNet consists of three modules: feature extraction, cost volume generation, and displacement map prediction. Firstly, the feature extraction module composed of the residual layer and the spatial pyramid pooling layer is designed to extract the features of different receptive fields. Then, an attention connect volume was proposed to represent the similarity information between pixels at different spatial scales. Finally, the displacement map is predicted by a stacked hourglass-shaped sub-net. Simulation and physical experiments verify that Stereo-DICNet can perform both temporal matching and stereo matching. Moreover, material tensile experiments show that the MAE of Stereo-DICNet in the U, V, and W directions are no more than 0.005 mm, 0.015 mm, 0.01 mm respectively.

Out-of-Plane Performance of Lightweight AAC Infills within RC Frames Using Noncontact Full-Field Strain Measurement Technique

Out-of-Plane Performance of Lightweight AAC Infills within RC Frames Using Noncontact Full-Field Strain Measurement Technique

Investigation of Nondestructive and Noncontact Electrical Characterization of GaN Thin Film on ScAlMgO4 Substrate Using Terahertz Time‐Domain Spectroscopic Ellipsometry with Characteristic Impedance Analytical Model

Semiconductor evaluation methods frequently require sample processing and carry the risk of defects and property changes. Therefore, nondestructive and noncontact electrical property measurement techniques are necessary. Terahertz time‐domain spectroscopic ellipsometry (THz‐TDSE) can simultaneously estimate the electrical properties and film thickness of a sample. However, whether this method can simultaneously determine electrical properties and film thickness for semiconductor thin films of 1 μm or less remains unclear. This study uses THz‐TDSE to evaluate the electrical properties and film thickness of a GaN thin film (≈0.8 μm thick) on a ScAlMgO4 substrate. Using the conventional analysis based on Fresnel's formula (FR model), which is used in the field of optics, uniquely determining the electrical properties and film thickness is not possible owing to multiple optimal solutions. Therefore, a model combining the characteristic impedance model used in the radio‐wave region with the FR model is employed in the analysis. The obtained values of electrical properties of the GaN thin film evaluated using THz‐TDSE are consistent with those using Hall effect measurement. Moreover, the film thickness is consistent with that measured from the cross‐sectional transmission electron microscopy observation.

Application of 3D Digital Image Correlation Technique to Measurement of Wind Blade Properties from Coupon Test and Small-Sized-Blade Frequency Test

As a wind blade is getting larger, its structural behavior becomes more complex and thus the risk of blade structural failure becomes higher. Generally, structural integrity is proved through blade testing with contact measurement devices such as strain gauges. However, contact measurement can only measure the physical quantities from installation points and requires lengthy cables to transmit the electrical signals to a data acquisition system, which can cause signal noise and add unnecessary weight. Recently, noncontact measurement techniques such as digital image correlation (DIC) have been widely used in various industries to process images captured from cameras and generate full-field strains and displacements without any mechanical connections such as cables. In this study, the application of three-dimensional DIC to wind blades was investigated; in addition, the material properties of glass-fiber-reinforced plastics used in blades for the measurement of the first frequency were also studied. Tensile testing was performed using coupons equipped with strain gauges and speckle patterns. In addition, a small blade from a 10 kW wind turbine was utilized to measure the first flapwise frequency using DIC and modal tests; the results showed a 2.1% difference between the two different measurements. This research showed that the DIC technique could be an alternative replacement for the traditional measurement technique upon further validation tests.

TSA and FEM Analysis Applied to CAD/CAM Titanium All-on-Four Prosthesis

Patient with fully edentulous maxillaries often suffer of several disease as an alteration of facial appearance, difficulties in eating, speech, and chews. Fixed and removable prosthesis can be a solution to rehabilitate this issue. The All-on-Four protocol is one of the possible treatment modalities for implant-supported prosthetics. It consists of placing four implants in an intermorainal position, two perpendicular to the occlusal plane and two tilted distally at a 45° angle. This paper aims to assess the stress distribution on an All-on-Four titanium prosthesis by means of a Thermoelastic Stress Analysis (TSA). This full-field non-contact measurement technique allows, through a thermal imaging camera, to observe how the surface temperature of an object changes under the effect of cyclic load. The principle on which thermoelastic analysis is based is the existence of a relationship between deformation, hence applied stress, and temperature change, called the thermoelastic effect. To perform the tests, the prosthesis was fixed to a resin jaw cast that has similar characteristics to the human bone, so that the same damping that would be present under normal working conditions could be accounted for. The distal artificial molar tooth was kept in contact to a narrow screw, which was attached to a Medium-Force (M-Series) Air-Cooled Shakers made by Santek. This shaker generates a sinusoidal load of 800N with a load cell at 10 Hz frequency, which can maintain the temperature variation of the prosthesis constant. A pre-load was applied before starting the tests. To see the change in temperature a Flir A6751SC thermal camera was used, and videos were acquired with a frame rate of 125 Hz and a resolution of 640x512 pixels. From the experimental TSA tests, the trend of the experimental stress concentration on the titanium specimen was detected. The outcomes of the experimental tests were compared to a 3D prosthesis model by Finite Element Analysis (FEM).

A mechanical model of ocular bulb vibrations and implications for acoustic tonometry.

In this study, we propose a comprehensive mechanical model of ocular bulb vibrations and discuss its implications for acoustic tonometry. The model describes the eye wall as a spherical, pre-stressed elastic shell containing a viscoelastic material and accounts for the interaction between the elastic corneoscleral shell and the viscoelastic vitreous humor. We investigate the natural frequencies of the system and the corresponding vibration modes, expanding the solution in terms of scalar and vector spherical harmonics. From a quantitative point of view, our findings reveal that the eyebulb vibration frequencies significantly depend on IOP. This dependency has two origins: "geometric" stiffening, due to an increase of the pre-stress, and "material" stiffening, due to the nonlinearity of the stress-strain curve of the sclera. The model shows that the second effect is by far dominant. We also find that the oscillation frequencies depend on ocular rigidity, but this dependency is important only at relatively large values of IOP. Thus close to physiological conditions, IOP is the main determinant of ocular vibration frequencies. The vitreous rheological properties are found to mostly influence vibration damping. This study contributes to the understanding of the mechanical behavior of the eye under dynamic conditions and thus has implications for non-contact intraocular pressure measurement techniques, such as acoustic tonometry. The model can also be relevant for other ocular pathological conditions, such as traumatic retinal detachment, which are believed to be influenced by the dynamic behavior of the eye.

Flood Discharge Measurement by STIV Coupled with Maximum Entropy Method Using Parameter Estimation by Numerical Analysis

The coupling of surface velocity measurement by image analysis and the maximum entropy method (MEM) is a non-contact measurement technique to estimate river discharge during flood with high accuracy. The MEM can represent complex vertical velocity distributions including velocity dips, but there is a challenge that the measurement of the cross-sectional velocity distribution by Acoustic Doppler Current Profiler (ADCP), etc. must be required at least once to determine the entropy parameter, which is a specific value at the cross section. This study proposes a new framework for determining the entropy parameter at new measurement sites based on the results of 3-D flow simulations. The framework is applied to two measurement sites, and the cross-sectional velocity distribution and the discharge estimated by the MEM with the entropy parameter are evaluated. The result shows that the framework enables the setting of an appropriate entropy parameter at a given site, and that the discharge estimated using the entropy parameter has a relative error of less than 10% to those measured by the ADCP.

Blade tip timing for multi-mode identification based on the blade vibration velocity

Blade Tip Timing (BTT) is a widely employed non-contact measurement technique for monitoring turbomachinery blade vibrations. The conventional BTT method typically involves computing the expected time of arrival (TOA) and subtracting it from the actual TOA to derive the time difference, subsequently used to calculate blade-tip displacement. However, calculating the expected TOA introduces computational errors. Moreover, obtaining blade-tip vibration displacement under variable rotational speeds necessitates pre-processing to remove trend terms caused by speed fluctuations. To address these challenges, this study introduces a paradigm shift by replacing the traditional displacement-based blade-tip timing (D-BTT) with velocity-based blade-tip timing (V-BTT). Specifically, this approach utilizes the angular information between adjacent sensors and the measured TOA to extract the blade-tip vibration velocity through a Taylor series expansion, thus eliminating the calculating of the expected TOA and improving the computational efficiency. Building upon this methodology, a compressed sensing model for undersampled vibration velocity is established and solved using the Block-OMP algorithm, enabling vibration parameter identification without relying on prior information. By providing a superior representation of mid-frequency components of blade-tip vibration, V-BTT facilitates precise identification of higher-order vibration parameters in multi-mode blade motion. Finally, the efficacy of the proposed method is confirmed through numerical simulations and experiments involving multi-mode blades. The experimental results reveal that the relative errors of the first two modal frequencies for the five blades identified using vibration velocity are all below 0.5%, affirming the method's accuracy. Compared to the traditional D-BTT method, the V-BTT approach reduces the need for trend removal, eliminates calculation errors associated with the expected TOA, and, most importantly, enhances the accuracy of multi-mode frequency identification.

Unsupervised CNN-based DIC method for 2D displacement measurement

Digital image correlation (DIC) is a widely used technique for non-contact measurement of deformation. However, traditional DIC methods face challenges in balancing calculation efficiency and the quantity of seed points. Deep learning approaches, particularly supervised learning methods, have shown promise in improving DIC efficiency. However, these methods require high-quality training data, which can be time-consuming to generate ground truth annotations. To address these challenges, we propose an unsupervised convolutional neural network (CNN) based DIC method for 2D displacement measurement. Our approach leverages an encoder-decoder architecture with multi-level feature extraction, a dual-path correlation block, and an attention block to extract informative features from speckle images with varying characteristics. We utilize a speckle image warp model to transform the deformed speckle image to the predicted reference speckle image based on the predicted 2D displacement map. The unsupervised training is achieved by comparing the predicted and original reference speckle images. To optimize the network's parameters, we employ a composite loss function that takes into account both the Mean Squared Error (MSE) and Pearson correlation coefficient. By using unsupervised convolutional neural network (CNN) based DIC method, we eliminate the need for extensive training data annotation, which is a time-consuming process in supervised learning DIC methods. We have conducted several experiments to demonstrate the validity and robustness of our proposed method. The results show a significant reduction in Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) compared to a method proposed by Zhao et al. This indicates that our unsupervised CNN-based DIC approach can achieve accuracy comparable to supervised CNN-based DIC methods. For implementation and evaluation, we provide PyTorch code and datasets, which will be released at the following URL :https://github.com/fead1/DICNet-corr-unsupervised-learning-.