19 June 2023 in IEEE Transactions on Instrumentation and Measurement
Structured light-based micro three-dimensional (3D) imaging technology is widely used in industrial applications. Using a well-calibrated camera and a projector, it is easy to obtain 3D surface topography of objects. However, single camera-projector pairs suffer from problems caused by shadows and their accuracy is limited due to the restricted field of view. To overcome these issues, this article proposes a multi-projector vision system consisting of a central camera and four surrounding projectors. The global geometry constraint scheme is applied to ensure consistency and stability across multiple projectors, which is achieved by board-to-board pose optimization. In addition, instead of directly using the iso-phase fringe line, geometric plane constraints are used in the calibration process to estimate corresponding projector pixel coordinates. Finally, a triangulation-based 3D reconstruction and data fusion approach is proposed to obtain high precision and shadow-reduced 3D surface topography. Experimental results are presented and discussed that demonstrate the good consistency and measurement accuracy of the proposed approach. The global geometry constraint scheme allows relative standard deviation and root-mean-square error in the measurements to be significantly reduced.
1 November 2021 in Neurocomputing
20 April 2022 in Applied Intelligence
12 May 2025 in Optics Letters
Deep learning techniques have been widely applied in fringe projection profilometry. However, existing methods focus on single-view measurements. This Letter introduces, for the first time to our knowledge, an end-to-end multi-view fringe projection profilometry (MVFPP) learning framework. It treats the decoding process as a deep feature correlation, enabling adaptation to a wide range of patterns and views of input using a feature transfer algorithm, which is guided by global spatial consistency constraints. Compared with traditional multi-stage MVFPP methods, our method more stably recovers dense surface representations, even under conditions of low coding bits, low reflectivity, and shadow occlusion. To better evaluate our method, a dataset of industrial electronics scenarios has been collected. Experimental results demonstrate that the proposed method can achieve state-of-the-art results with flexible coding patterns and coding bits.
23 May 2025 in IEEE Transactions on Instrumentation and Measurement
Structured light illumination is a remarkable technique for active three-dimensional measurements. However, global or indirect light transport degrades the accuracy of phase calculation. This article describes an efficient phase correction and fusion framework for microscopic measurements. It targets a multi-projector system consisting of one camera coupled with multiple digital projectors. The phase-to-depth relationship is determined by using the native geometric model, taking lens distortion into consideration. By exploiting spatial consistency across multiple views, the phase from different projectors can be transformed into a unified view space. In this context, a reliability metric for unified phases is proposed to identify erroneous phase regions. Using preserved consistent views and the weighted histogram of the local neighborhood, phase error is first coarsely compensated by rectifying the doubtful period order. In addition, a novel hybrid weighting factor integrating signal and geometric reliability is proposed to further reduce measurement error. Experimental results from both qualitative and quantitative evaluation demonstrate the feasibility and usefulness of the proposed framework.
12 June 2025 in IEEE Transactions on Industrial Informatics
Structured light imaging is a typical technique for industrial 3-D microscopic measurement. Extensive research on structured light codecs has been conducted to accurately correlate camera and projector pixels. However, these methods suffer significant degradation when measuring low-reflectivity and complex surfaces. This article introduces a deep correlation-based cascade structured light network (CasSLNet) that utilizes deep phase and column features to calculate correspondences at the subpixel scale. To mitigate the huge computational cost of full correlation, a coarse-to-fine approach is proposed. Specifically, multiscale features from the camera observation sequence and the 1-D encoding pattern are extracted through a pseudosiamese network, and cascade cost volumes are constructed. An initial column map is then regressed from the low-resolution column cost volume. Based on this, an iterative update operator is introduced to refine initial estimates, resulting in a full-resolution column map. Furthermore, a structured light dataset has been collected and experiments have been conducted on a typical structured light imaging platform. Experimental results demonstrate that CasSLNet outperforms both traditional and state-of-the-art deep learning-based methods.