Geometric Calibration of Micro-Lens-Based Light Field Cameras Using Line Features

Bok, Yunsu; Jeon, Hae‐Gon; Kweon, In So

doi:10.1109/tpami.2016.2541145

Cited by 147 publications

(82 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In terms of the grid accuracy Q g , we find that the proposed method yields results of about one order of magnitude better than the other methods while occasionally showing extreme results performing even two orders of magnitude better. We observe that the grid estimation proposed by Bok et al [4] slightly outperforms the one by Dansereau et al [7]. Even though the proposed method has a longer runtime, with an average of about 160 s per WI compared to an average we argue that this is still feasible, as the calibration usually only has to be executed once per camera.…”

Section: Resultsmentioning

confidence: 68%

“…Note that, since the calibration by Bok at al. [4] does not utilize a regular grid but the individually detected centers, the grid spacing and rotation accuracies Q s , Q r cannot be specified in this case. We observe that only the proposed algorithm satisfies the accuracy requirements as stated in Section III-A.…”

Section: Resultsmentioning

confidence: 99%

“…Usually, the ML grid is estimated by detecting the ML centers from the corresponding WI [4] and building a regular grid best approximating the detected centers [7] or by directly estimating regular grid from the WI [6]. Challenges in the detection are versatile: on the one hand, the sheer amount of MLs in MLAs used in practice limits the algorithm's complexity.…”

Section: Microlens Grid Estimationmentioning

confidence: 99%

“…But since it does not yield results that are significantly different from those of Dansereau et al [7], we do not re-evaluate it here. For completeness, we will also evaluate the ML center detection used by Bok et al [4], which is not presented in their paper but implemented in the corresponding MATLAB reference implementation. None of the mentioned algorithms consider vignetting effects.…”

Section: Microlens Grid Estimationmentioning

confidence: 99%

“…by using the ray re-projection error. In most light field camera models [7], [6], [22], [23], the calibration is performed using image-side light fields that have been decoded from the sensor lenslet images, while others perform calibration utilizing the raw images directly [4]. In either case, this includes multiple non-trivial pre-processing steps, such as the detection of the projected microlens (ML) centers and estimation of a regular grid approximating the centers, alignment of the lenslet image with the sensor, slicing the image into a light field and, in the case of hexagonal MLAs, resampling the light field onto a rectangular grid.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Microlens Array Grid Estimation, Light Field Decoding, and Calibration

Schambach

León

2020

IEEE Trans. Comput. Imaging

View full text Add to dashboard Cite

We quantitatively investigate multiple algorithms for microlens array grid estimation for microlens array-based light field cameras. Explicitly taking into account natural and mechanical vignetting effects, we propose a new method for microlens array grid estimation that outperforms the ones previously discussed in the literature. To quantify the performance of the algorithms, we propose an evaluation pipeline utilizing application-specific ray-traced white images with known microlens positions. Using a large dataset of synthesized white images, we thoroughly compare the performance of the different estimation algorithms. As an example, we apply our results to the decoding and calibration of light fields taken with a Lytro Illum camera. We observe that decoding as well as calibration benefit from a more accurate, vignetting-aware grid estimation, especially in peripheral subapertures of the light field.Since the ML grid parameters are application-specific, we make the source code for a full evaluation pipeline (image

show abstract

Section: Resultsmentioning

confidence: 68%

Section: Resultsmentioning

confidence: 99%

Section: Microlens Grid Estimationmentioning

confidence: 99%

Section: Microlens Grid Estimationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Microlens Array Grid Estimation, Light Field Decoding, and Calibration

Schambach

León

2020

IEEE Trans. Comput. Imaging

View full text Add to dashboard Cite

show abstract

A Meta‐Device for Intelligent Depth Perception

Chen

Liu

et al. 2022

Advanced Materials

View full text Add to dashboard Cite

The optical illusion affects depth‐sensing due to the limited and specific light‐field information acquired by single‐lens imaging. The incomplete depth information or visual deception would cause cognitive errors. To resolve this problem, an intelligent and compact depth‐sensing meta‐device that is miniaturized, integrated, and applicable for diverse scenes in all light levels is demonstrated. The compact and multifunction stereo vision system adopts an array with 3600 achromatic meta‐lenses and a size of 1.2 × 1.2 mm2 to measure the depth over a 30 cm range with deep‐learning support. The meta‐lens array can act as multiple imaging lenses to collect light field information. It can also work with a light source as an active optical device to project a structured light. The meta‐lens array can serve as the core functional component of a light‐field imaging system under bright conditions or a structured‐light projection system in the dark. The depth information in both ways can be analyzed and extracted by the convolutional neural network. This work provides a new avenue for the applications such as autonomous driving, machine vision, human–computer interaction, augmented reality, biometric identification, etc.

show abstract

Machine‐Learned Light‐Field Camera that Reads Facial Expression from High‐Contrast and Illumination Invariant 3D Facial Images

Bae

Lee

Kwon

et al. 2021

Advanced Intelligent Systems

View full text Add to dashboard Cite

Facial expression conveys nonverbal communication information to help humans better perceive physical or psychophysical situations. Accurate 3D imaging provides stable topographic changes for reading facial expression. In particular, light‐field cameras (LFCs) have high potential for constructing depth maps, thanks to a simple configuration of microlens arrays and an objective lens. Herein, machine‐learned NIR‐based LFCs (NIR‐LFCs) for facial expression reading by extracting Euclidean distances of 3D facial landmarks in pairwise fashion are reported. The NIR‐LFC contains microlens arrays with asymmetric Fabry−Perot filter and NIR bandpass filter on CMOS image sensor, fully packaged with two vertical‐cavity surface‐emitting lasers. The NIR‐LFC not only increases the image contrast by 2.1 times compared with conventional LFCs, but also reduces the reconstruction errors by up to 54%, regardless of ambient illumination conditions. A multilayer perceptron (MLP) classifies input vectors, consisting of 78 pairwise distances on the facial depth map of happiness, anger, sadness, and disgust, and also exhibits exceptional average accuracy of 0.85 (p<0.05). This LFC provides a new platform for quantitatively labeling facial expression and emotion in point‐of‐care biomedical, social perception, or human−machine interaction applications.

show abstract

Geometric Calibration of Micro-Lens-Based Light Field Cameras Using Line Features

Cited by 147 publications

References 22 publications

Microlens Array Grid Estimation, Light Field Decoding, and Calibration

Microlens Array Grid Estimation, Light Field Decoding, and Calibration

A Meta‐Device for Intelligent Depth Perception

Machine‐Learned Light‐Field Camera that Reads Facial Expression from High‐Contrast and Illumination Invariant 3D Facial Images

Contact Info

Product

Resources

About