Large depth range binary-focusing projection 3D shape reconstruction via unpaired data learning

Optics and Lasers in Engineering, Journal Year: 2024, Volume and Issue: 181, P. 108442 - 108442

Published: July 22, 2024

Language: Английский

---

Bayesian-neural-network-based strain estimation approach for optical coherence elastography

Optica, Journal Year: 2024, Volume and Issue: 11(9), P. 1334 - 1334

Published: Sept. 5, 2024

Strain estimation is critical for quantitative elastography in quasi-static phase-sensitive optical coherence (PhS-OCE). Deep-learning methods have achieved exceptional performance estimating high-quality strain distributions. However, they cannot often assess their predictive accuracy and reliability rigorously. To navigate these challenges, a Bayesian-neural-network (BNN)-based proposed. The method can provide the uncertainty distribution of results beyond achieving estimation. Such an results. Moreover, degree function as indicator compensating phase decorrelation thus significantly enhancing SNR dynamic range PhS-OCE. Thermal three-point bending deformation experiments validated that predicted effectively address allow more comprehensive understanding estimated

Language: Английский

---

Synchronous edge-enhanced and bright-field 3D imaging in single-shot FINCH enabled by deep learning

Optics and Lasers in Engineering, Journal Year: 2025, Volume and Issue: 186, P. 108824 - 108824

Published: Jan. 7, 2025

Language: Английский

---

Unsupervised crosstalk suppression for self-interference digital holography

Optics Letters, Journal Year: 2025, Volume and Issue: 50(4), P. 1261 - 1261

Published: Jan. 22, 2025

Self-interference digital holography extends the application of to non-coherent imaging fields such as fluorescence and scattered light, providing a new solution, best our knowledge, for wide field 3D low coherence or partially coherent signals. However, cross talk information has always been an important factor limiting resolution this method. The suppression is complex nonlinear problem, deep learning can easily obtain its corresponding model through data-driven methods. in real experiments, it difficult paired datasets complete training. Here, we propose unsupervised method based on cycle-consistent generative adversarial network (CycleGAN) self-interference holography. Through introduction saliency constraint, model, named crosstalk suppressing with neural (CS-UNN), learn mapping between two image domains without requiring training data while avoiding distortions content. Experimental analysis shown that suppress reconstructed images need strategies large number datasets, effective solution technology.

Language: Английский

---

Adaptive few-shot image augmentation for fine-grained industrial defects based on region-level modeling

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 152, P. 110695 - 110695

Published: April 14, 2025

Language: Английский

---

Cross-Scale Hybrid Gaussian Attention Network for Object Detection in Remote Sensing Images

IEEE Geoscience and Remote Sensing Letters, Journal Year: 2024, Volume and Issue: 21, P. 1 - 5

Published: Jan. 1, 2024

Accurate object detection in remote sensing images (RSIs) is of great significance for various applications such as environmental monitoring and agricultural production. However, it a challenging task mainly due to the complex backgrounds scale diversity geospatial objects. In this letter, Cross-Scale Hybrid Gaussian Attention Network (CSHGANet) proposed accurate RSIs, consists two main components follows. First, hybrid attention designed learn interrelationships between channels spatial locations features, which can focus on objects reduce interference RSIs. Then, cross-scale feature aggregation module developed adaptively fuse multi-scale maps capture more rich discriminative representations, so better handle variations Extensive experiments public datasets (i.e., NWPU VHR-10 RSOD) show that CSHGANet outperforms state-of-the-art methods, achieving mean average precision (mAP) scores 95.53% 98.61%, respectively.

Language: Английский

---

Temporal Fusion Dynamically Separable Graph Convolutional Network for EEG Motion Intention Decoding Based on Source Information Extraction

IEEE Transactions on Instrumentation and Measurement, Journal Year: 2024, Volume and Issue: 73, P. 1 - 13

Published: Jan. 1, 2024

Language: Английский

---

Handheld structured light system for panoramic 3D measurement in mesoscale

Measurement Science and Technology, Journal Year: 2024, Volume and Issue: 35(10), P. 105015 - 105015

Published: July 12, 2024

Abstract The measurement of complete 3D topography in mesoscale plays a vital role high-precision reverse engineering, oral medical modeling, circuit detection, etc. Traditional structured light systems are limited to measuring shapes from single perspective. Achieving high-quality mesoscopic panoramic remains challenging, especially complex measured scenarios such as dynamic measurement, scattering mediums, and high reflectance. To overcome these problems, we develop handheld system for scenes together with the fast point-cloud-registration accurate 3D-reconstruction, where motion discrimination mechanism is designed ensure that captured fringe quasi-stationary case by avoiding errors caused during scanning; deep neural network utilized suppress degradation resulting significant improvement quality point cloud; strategy based on phase averaging additionally proposed simultaneously correct saturation-induced gamma nonlinear errors. Finally, incorporates multi-threaded data processing framework verify method, corresponding experiments its feasibility.

Language: Английский

---

Phase volume correlation approach for overcoming decorrelation in three-dimensional phase-sensitive optical coherence elastography

Optics Express, Journal Year: 2024, Volume and Issue: 32(22), P. 38437 - 38437

Published: Sept. 30, 2024

The dynamic measurement range in phase-sensitive optical coherence elastography (PhS-OCE) is limited for the phase decorrelation induced by pixel-level displacements precision measurement, where consideration of time-resolved incremental method and in-plane pixels tracking insufficient to recover holistically. This work presented a volume correlation (PVC) approach handle three-dimensional PhS-OCE. By utilizing ability discontinuous source diagram quantify voxel levels, PVC establishes wrapped phase-matching equation aimed at optimizing number volumetric distributions. motions deformed space can be evaluated solving optimization model matching, thereby enabling reconstruction variation corrupted decorrelation. large deformations experiments including diffident loadings, i.e., stretching, three-point bending, light-cured, verified proposed PPVC approach's feasibility, reliability, stability. contribution this dramatically enhance measuring

Language: Английский

---

Suppression of Defocused Images in Digital Holographic Reconstruction Based on Image Plane Filtering Technique

Frontiers in Optics + Laser Science 2022 (FIO, LS), Journal Year: 2024, Volume and Issue: unknown, P. JD4A.91 - JD4A.91

Published: Jan. 1, 2024

We use image plane filtering technique to filter out the focused light field in digital holographic reconstruction, which can reduce defocusing effect of this part on reconstruction at other distances.

Language: Английский

---