Three-dimensional (3D) imaging with structured light is crucial in diverse scenarios, ranging from intelligent manufacturing and medicine to entertainment. However, current structured light methods rely on projector–camera synchronization, limiting the use of affordable imaging devices and their consumer applications. In this work, we introduce an asynchronous structured light imaging approach based on generative deep neural networks to relax the synchronization constraint, accomplishing the challenges of fringe pattern aliasing, without relying on any a priori constraint of the projection system. To overcome this need, we propose a generative deep neural network with U-Net-like encoder–decoder architecture to learn the underlying fringe features directly by exploring the intrinsic prior principles in the fringe pattern aliasing. We train within an adversarial learning framework and supervise the network training via a statistics-informed loss function. We demonstrate that by evaluating the performance on fields of intensity, phase, and 3D reconstruction. It is shown that the trained network can separate aliased fringe patterns for producing comparable results with the synchronous one: the absolute error is no greater than 8 μm, and the standard deviation does not exceed 3 μm. Evaluation results on multiple objects and pattern types show it could be generalized for any asynchronous structured light scene.
KEYWORDS: Calibration, Cameras, Distortion, 3D modeling, 3D acquisition, Stereoscopic cameras, 3D metrology, Visualization, Optical engineering, Visual process modeling
Camera calibration is crucial for geometric vision-based three-dimensional (3D) deformation measurement tasks. Among existing calibration techniques, the one based on planar targets has attracted much attention in the community due to its flexibility and reliability. Our study proposes a calibration technique to obtain high-accuracy internal and external parameters based on low-cost ordinary planar patterns. The proposed method determines the optimal internal parameters for each camera by refining 3D coordinates of planar control points, where an analytic model of optics distortion is presented to enable lens distortion to be corrected directly in subsequent external calibration and underlying 3D reconstruction. External parameters are estimated by minimizing a bundle adjustment framework, which is carefully designed based on the proposed distortion correction model and depth parameterization. In contrast to the existing techniques, the proposed method is capable of obtaining a high-accuracy calibration with ordinary targets rather than the well-designed and fabricated ones. We experimented the proposed method with a calibration performance analysis and a displacement measurement; both results demonstrated the accuracy and robustness.
Object detection in aerial images plays an important role for a wide range of applications. Although many efforts have been done in the last decade, it is still an active and challenging problem because of the highly complex backgrounds and the large variations in the visual appearance of objects caused by viewpoint variation, occlusion, illumination, etc. Recently, many object detectors based on deep learning demonstrate the great advantages for significantly improving the detection performance in aerial images. However, the most accuracy neural networks usually have hundreds of layers and thousands of channels, thus requiring huge computation and memory consumption. Besides, the state-of-the-art object detectors are usually fined-tuned from the models pretrained on classification dataset ImageNet, which limits the modification of network architecture and also leads to learning bias because of the different domains. In this paper we trained a lightweight convolutional neural network from scratch to perform object detection in aerial images. When designing the lightweight network, Concatenated Rectified Linear Units (CReLU) and depthwise separable convolution operation were employed to reduce the computation cost and model size. When training the lightweight network from scratch, we employ Group Normalization (GN) in each convolution layer, which makes smoother optimization landscape and has more stable gradients. A serial of ablation experiments is conducted on the recently published large-scale Dataset for Object detection in Aerial images (DOTA), and the results show that the proposed object detection methods with lightweight network trained from scratch achieves competitive performance but has smaller model size and lower computation cost.
Videometrics is a technique for measuring displacement, deformation and motion with features of precision, multifunctional, automation and real time measurement etc. Videometrics with camera networks is a fast developed area for deformation measurements of large scale structures. Conventional camera network is parallel network where cameras are independent each other and the relations among the cameras are calibrated from their target images. In recent years, we proposed and developed two kinds of videometrics with camera series networks where cameras are connected each other in series and relations among the cameras can be relayed one by one for the deformation measurements of large and super large scale structures. In this paper, our research work in both the camera series and parallel networks for the deformation measurements of large scale structures are overviewed and some new development are introduced. First, our proposed methods of camera series networks are introduced, including the pose-relay videometrics with camera series and the displacement-relay videometrics with camera series. Then our work of large scale structure deformation measurement by camera parallel networks is overviewed. Videometrics with various types of camera networks has the broad prospect of undertaking automatic, long-term and continuous measurement for deformation in engineering projects such as wind turbine blades, ship, railroad beds, and bridges.
A simple and flexible method for non-overlapping camera rig calibration that includes camera calibration and relative poses calibration is presented. The proposed algorithm gives the solutions of the cameras parameters and the relative poses simultaneously by using nonlinear optimization. Firstly, the intrinsic and extrinsic parameters of each camera in the rig are estimated individually. Then, a linear solution derived from hand-eye calibration scheme is proposed to compute an initial estimate of the relative poses inside the camera rig. Finally, combined non-linear refinement of all parameters is performed, which optimizes the intrinsic parameters, the extrinsic parameters and relative poses of the coupled camera at the same time. We develop and test a novel approach for calibrating the parameters of non-overlapping camera rig using camera calibration and hand-eye calibration method. The method is designed inter alia for the purpose of deformation measurement using the calibrated rig. Compared the camera calibration with hand-eye calibration separately, our joint calibration is more convenient in practice application. Experimental data shows our algorithm is feasible and effective.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.