A5077125791- Optical measurement and interference techniques
- Image Processing Techniques and Applications
- Advanced Vision and Imaging
- Advanced Measurement and Metrology Techniques
- Digital Holography and Microscopy
- Advanced Image Processing Techniques
- Semiconductor Lasers and Optical Devices
- Water Quality Monitoring and Analysis
- Advanced Fluorescence Microscopy Techniques
- Advanced Image Fusion Techniques
- Robotics and Sensor-Based Localization
- Retinal Imaging and Analysis
- Optical Coherence Tomography Applications
- Advanced Optical Imaging Technologies
- Advanced optical system design
- Photonic and Optical Devices
Huawei Technologies (United States)
2025
Université Claude Bernard Lyon 1
2024
Purdue University West Lafayette
2021-2024
École Centrale de Lyon
2006
Abstract: Miniaturized pixel sizes in near-eye digital displays lead to emission patterns with large divergence angles, necessitating efficient beam collimation solutions improve the light coupling efficiency. Traditional optics, such as lenses and cavities, are wavelength-sensitive cannot simultaneously collimate red (R), green (G), blue (B) light. In this work, we employed inverse design optimization finite-difference time-domain (FDTD) simulation techniques a collimator comprised of...
This paper presents a calibration method for microscopic structured light system with an extended depth of field (DOF). We first employed the focal sweep technique to achieve large enough measurement range, and then developed computational framework alleviate impact phase errors caused by standard off-the-shelf target (black circles white background). Specifically, we polynomial interpolation algorithm correct near black obtain more accurate maps projector feature points determination....
This Letter presents a novel, to the best of our knowledge, method calibrate multi-focus microscopic structured-light three-dimensional (3D) imaging systems with an electrically adjustable camera focal length. We first leverage conventional system reference length f0. Then we other discrete lengths fi by determining virtual features on reconstructed white plane using Finally, fit polynomial function model calibration results for fi. Experimental demonstrate that proposed can consistently and...
In this paper, we develop a new hybrid active vision/geometric modeling approach dedicated to 3D human face recovery. Initially, coarse reconstruction is obtained via structured-light assisted stereo sensor. Here, the matching problem resolved through two main stages: (a) sub-pixel stripe edge localization of projected patterns light, and (b) correspondences establishing based on adaptive dynamic programming optimization technique. Next, introduction smooth interpolation models achieves fine...
This paper presents an optimized capturing strategy for large depth-of-field (DOF) 3D microscopic structured-light imaging with the focus stacking technique. Different from conventional method that captures fringe images under a series of pre-defined settings, proposed automatically determines effective settings measured objects and only these settings. Specifically, first roughly measures depths using focal sweep technique, then calculates locate camera plane on valid objects. Finally,...
This paper presents an adaptive focus stacking method for large depth-of-field (DOF) 3D microscopic structured-light imaging systems. Conventional methods typically capture images under a series of pre-defined settings without considering the attributes measured object. Therefore, it is inefficient since some might be redundant. To address this problem, we first employ focal sweep technique to reconstruct initial rough shape objects. Then, leverage data determine effective that camera on...
Retinal image quality assessment (RIQA) is crucial for diagnosing various eye diseases and ensuring the accuracy of diagnostic analyses based on retinal fundus images. Traditional deep convolutional neural networks (CNNs) RIQA face challenges such as over-reliance RGB brightness difficulty in differentiating closely ranked categories. To address these issues, we introduced Dual-Path Frequency-domain Cross-attention Network (DFC-Net), which integrates images contrast-enhanced using...
3D face modeling is a very important and challenging task in computer vision especially the presence of an expression variation.It plays role wide range applications including facial animation, human-computer interaction surgery.Current imaging solutions are often based on structured light camera/projector systems to overcome relatively uniform appearance skin.This paper proposes space-time scheme for scanning, employing three calibrated cameras coupled with non projector device.The proposed...
This paper presents a novel phase unwrapping algorithm via depth from focus method in microscopic fringe projection profilometry. The proposed uses contrast to estimate the rough information and determines orders by geometrical constraint relationships. As result, it does not require extra patterns or images, enabling higher 3D imaging speed. Experimental results demonstrate that can successfully realize