Cameras are a crucial exteroceptive sensor for self-driving cars as they are low-cost and small, provide appearance information about the environment, and work in various weather conditions. They can be used for multiple purposes such as visual navigation and obstacle detection. We can use a surround multi-camera system to cover the full 360-degree field-of-view around the car. In this way, we avoid blind spots which can otherwise lead to accidents. To minimize the number of cameras needed for surround perception, we utilize fisheye cameras. Consequently, standard vision pipelines for 3D mapping, visual localization, obstacle detection, etc. need to be adapted to take full advantage of the availability of multiple cameras rather than treat each camera individually. In addition, processing of fisheye images has to be supported. In this paper, we describe the camera calibration and subsequent processing pipeline for multi-fisheye-camera systems developed as part of the V-Charge project. This project seeks to enable automated valet parking for self-driving cars. Our pipeline is able to precisely calibrate multi-camera systems, build sparse 3D maps for visual navigation, visually localize the car with respect to these maps, generate accurate dense maps, as well as detect obstacles based on real-time depth map extraction.
Depth sensing devices have created various new applications in scientific and commercial research with the advent of Microsoft Kinect and PMD (Photon Mixing Device) cameras. Most of these applications require the depth cameras to be pre-calibrated. However, traditional calibration methods using a checkerboard do not work very well for depth cameras due to the low image resolution. In this paper, we propose a depth calibration scheme which excels in estimating camera calibration parameters when only a handful of corners and calibration images are available. We exploit the noise properties of PMD devices to denoise depth measurements and perform camera calibration using the denoised depth as an additional set of measurements. Our synthetic and real experiments show that our depth denoising and depth based calibration scheme provides significantly better results than traditional calibration methods.
Recent direct visual odometry and SLAM algorithms have demonstrated impressive levels of precision. However, they require a photometric camera calibration in order to achieve competitive results. Hence, the respective algorithm cannot be directly applied to an off-the-shelf-camera or to a video sequence acquired with an unknown camera. In this work we propose a method for online photometric calibration which enables to process auto exposure videos with visual odometry precisions that are on par with those of photometrically calibrated videos. Our algorithm recovers the exposure times of consecutive frames, the camera response function, and the attenuation factors of the sensor irradiance due to vignetting. Gain robust KLT feature tracks are used to obtain scene point correspondences as input to a nonlinear optimization framework. We show that our approach can reliably calibrate arbitrary video sequences by evaluating it on datasets for which full photometric ground truth is available. We further show that our calibration can improve the performance of a state-of-the-art direct visual odometry method that works solely on pixel intensities, calibrating for photometric parameters in an online fashion in realtime.
We present a novel approach for vanishing point detection from uncalibrated monocular images. In contrast to state-of-the-art, we make no a priori assumptions about the observed scene. Our method is based on a convolutional neural network (CNN) which does not use natural images, but a Gaussian sphere representation arising from an inverse gnomonic projection of lines detected in an image. This allows us to rely on synthetic data for training, eliminating the need for labelled images. Our method achieves competitive performance on three horizon estimation benchmark datasets. We further highlight some additional use cases for which our vanishing point detection algorithm can be used.
We present a method for the real-time estimation of the full 3D pose of one or more human hands using a single commodity RGB camera. Recent work in the area has displayed impressive progress using RGBD input. However, since the introduction of RGBD sensors, there has been little progress for the case of monocular color input. We capitalize on the latest advancements of deep learning, combining them with the power of generative hand pose estimation techniques to achieve real-time monocular 3D hand pose estimation in unrestricted scenarios. More specifically, given an RGB image and the relevant camera calibration information, we employ a state-of-the-art detector to localize hands. Given a crop of a hand in the image, we run the pretrained network of OpenPose for hands to estimate the 2D location of hand joints. Finally, non-linear least-squares minimization fits a 3D model of the hand to the estimated 2D joint positions, recovering the 3D hand pose. Extensive experimental results provide comparison to the state of the art as well as qualitative assessment of the method in the wild.
In this paper we present a new camera calibration method aimed at finding a straight-line locus, in a special colour feature space, that is traversed by daylights and as well also approximately followed by specular points. The aim of the calibration is to enable recovering the colour of the illuminant in a scene, using the calibrated camera. First we prove theoretically that any candidate specular points, for an image that is generated by a specific camera and taken under a daylight, must lie on a straight line in log-chromaticity space, for a chromaticity that is generated using a geometric-mean denominator. Use is made of the assumptions that daylight illuminants can be approximated using Planckians and that camera sensors are narrowband or can be made so by spectral sharpening. Then we show how a particular camera can be calibrated so as to discover this locus. As applications we use this curve for illuminant detection, and also for re-lighting of images to show they would appear under lighting having a different colour temperature.
Calibrating narrow field of view soccer cameras is challenging because there are very few field markings in the image. Unlike previous solutions, we propose a two-point method, which requires only two point correspondences given the prior knowledge of base location and orientation of a pan-tilt-zoom (PTZ) camera. We deploy this new calibration method to annotate pan-tilt-zoom data from soccer videos. The collected data are used as references for new images. We also propose a fast random forest method to predict pan-tilt angles without image-to-image feature matching, leading to an efficient calibration method for new images. We demonstrate our system on synthetic data and two real soccer datasets. Our two-point approach achieves superior performance over the state-of-the-art method.
3D LiDARs and 2D cameras are increasingly being used alongside each other in sensor rigs for perception tasks. Before these sensors can be used to gather meaningful data, however, their extrinsics (and intrinsics) need to be accurately calibrated, as the performance of the sensor rig is extremely sensitive to these calibration parameters. A vast majority of existing calibration techniques require significant amounts of data and/or calibration targets and human effort, severely impacting their applicability in large-scale production systems. We address this gap with CalibNet: a self-supervised deep network capable of automatically estimating the 6-DoF rigid body transformation between a 3D LiDAR and a 2D camera in real-time. CalibNet alleviates the need for calibration targets, thereby resulting in significant savings in calibration efforts. During training, the network only takes as input a LiDAR point cloud, the corresponding monocular image, and the camera calibration matrix K. At train time, we do not impose direct supervision (i.e., we do not directly regress to the calibration parameters, for example). Instead, we train the network to predict calibration parameters that maximize the geometric and photometric consistency of the input images and point clouds. CalibNet learns to iteratively solve the underlying geometric problem and accurately predicts extrinsic calibration parameters for a wide range of mis-calibrations, without requiring retraining or domain adaptation. The project page is hosted at https://epiception.github.io/CalibNet
In this paper, we propose a method of targetless and automatic Camera-LiDAR calibration. Our approach is an extension of hand-eye calibration framework to 2D-3D calibration. By using the sensor fusion odometry method, the scaled camera motions are calculated with high accuracy. In addition to this, we clarify the suitable motion for this calibration method. The proposed method only requires the three-dimensional point cloud and the camera image and does not need other information such as reflectance of LiDAR and to give initial extrinsic parameter. In the experiments, we demonstrate our method using several sensor configurations in indoor and outdoor scenes to verify the effectiveness. The accuracy of our method achieves more than other comparable state-of-the-art methods.
Most current single image camera calibration methods rely on specific image features or user input, and cannot be applied to natural images captured in uncontrolled settings. We propose directly inferring camera calibration parameters from a single image using a deep convolutional neural network. This network is trained using automatically generated samples from a large-scale panorama dataset, and considerably outperforms other methods, including recent deep learning-based approaches, in terms of standard L2 error. However, we argue that in many cases it is more important to consider how humans perceive errors in camera estimation. To this end, we conduct a large-scale human perception study where we ask users to judge the realism of 3D objects composited with and without ground truth camera calibration. Based on this study, we develop a new perceptual measure for camera calibration, and demonstrate that our deep calibration network outperforms other methods on this measure. Finally, we demonstrate the use of our calibration network for a number of applications including virtual object insertion, image retrieval and compositing.
In recent years, many publications showed that convolutional neural network based features can have a superior performance to engineered features. However, not much effort was taken so far to extract local features efficiently for a whole image. In this paper, we present an approach to compute patch-based local feature descriptors efficiently in presence of pooling and striding layers for whole images at once. Our approach is generic and can be applied to nearly all existing network architectures. This includes networks for all local feature extraction tasks like camera calibration, Patchmatching, optical flow estimation and stereo matching. In addition, our approach can be applied to other patch-based approaches like sliding window object detection and recognition. We complete our paper with a speed benchmark of popular CNN based feature extraction approaches applied on a whole image, with and without our speedup, and example code (for Torch) that shows how an arbitrary CNN architecture can be easily converted by our approach.
In this paper, we focus on traffic camera calibration and a visual speed measurement from a single monocular camera, which is an important task of visual traffic surveillance. Existing methods addressing this problem are difficult to compare due to a lack of a common data set with reliable ground truth. Therefore, it is not clear how the methods compare in various aspects and what factors are affecting their performance. We captured a new data set of 18 full-HD videos, each around 1 hr long, captured at six different locations. Vehicles in the videos (20865 instances in total) are annotated with the precise speed measurements from optical gates using LiDAR and verified with several reference GPS tracks. We made the data set available for download and it contains the videos and metadata (calibration, lengths of features in image, annotations, and so on) for future comparison and evaluation. Camera calibration is the most crucial part of the speed measurement; therefore, we provide a brief overview of the methods and analyze a recently published method for fully automatic camera calibration and vehicle speed measurement and report the results on this data set in detail.
Camera calibration for estimating the intrinsic parameters and lens distortion is a prerequisite for various monocular vision applications including feature tracking and video stabilization. This application paper proposes a model for estimating the parameters on the fly by fusing gyroscope and camera data, both readily available in modern day smartphones. The model is based on joint estimation of visual feature positions, camera parameters, and the camera pose, the movement of which is assumed to follow the movement predicted by the gyroscope. Our model assumes the camera movement to be free, but continuous and differentiable, and individual features are assumed to stay stationary. The estimation is performed online using an extended Kalman filter, and it is shown to outperform existing methods in robustness and insensitivity to initialization. We demonstrate the method using simulated data and empirical data from an iPad.
카메라를 장착한 이동 로봇을 위한 레이저 트래커 보조 핸드-아이(hand-eye) 캘리브레이션 방법을 제안한다. 본 방법은 레이저 트래커 기반 3D 계측(metrology)과 카메라 기반 2D 관측을 결합한다. 지상 관측 이동 로봇을 대상으로 한 선행 연구의 레이저 트래커 보조 카메라-로봇 캘리브레이션 방법을 바탕으로, 트래커로 위치가 특정된 이동 로봇 좌표계에서 카메라 자세(pose)를 보정하기 위한 일반화된 정식화를 제시한다. 새 접근법은 여러 캘리브레이션 타깃을 연쇄(chain)함으로써 선행 방법이 로봇 및 카메라 구성에 대해 두었던 가정을 완화하고, 다양한 카메라 장착 이동 로봇 시스템을 지원하는 보다 일반적인 방법을 제공한다.
지능형 교통 시스템에서 도로변 3D 객체 검출은 교통 이해, 협력형 조기 경보, 안전한 자율주행에 필수적인 넓은 영역 인식을 제공한다. 그러나 기존 방법들은 카메라 외부 파라미터(extrinsic)에 대한 높은 민감도로 인한 문제가 있다. 일시적인 지터(jitter)나 지속적인 드리프트(drift)로 나타나는 미세한 편차조차 투영 기하학에 의해 크게 증폭될 수 있다. 이러한 연쇄 효과는 심각한 특징 불일치와 저하된 위치 추정을 초래한다. 이 한계를 완화하기 위해, 우리는 조각별 6-DoF 자세 오프셋을 사용하여 외부 파라미터를 보정하는 영역 인식 외부 파라미터 보정 프레임워크인 RECO를 제안한다. RECO는 학습 가능한 거리 경계를 예측하여 장면을 근거리와 원거리 영역으로 분할하고, 영역별 자세 보정을 추정한다. 미분 가능한 시그모이드 게이트가 두 보정된 기하를 부드럽게 혼합하여 연속적인 BEV 샘플링을 유지하고 안정적인 최적화를 촉진한다. 외부 파라미터 정제를 지도하기 위해, 3D ground truth에서 투영된 2D 바운딩 박스와 2D 어노테이션을 비교하는 보조 재투영 손실을 도입하고, 이를 표준 검출 목적 함수와 함께 공동 최적화한다. 외부 파라미터 섭동 하에서 DAIR-V2X-I 및 Rope3D 벤치마크에 대한 광범위한 실험은 yaw 및 $z$축 편차 모두에서 최신(state-of-the-art) 기준선 대비 일관된 성능 향상을 보여준다. RECO는 일시적 섭동에서 지속적 이동으로도 일반화되며, 엄격한 보정(calibration) 불확실성 하에서도 매우 경쟁력 있는 성능을 유지한다.
마커리스 모션 추적(Markerless motion tracking)은 지난 10년간 빠르게 발전해왔으며, 현재 행동, 임상, 생체역학 연구에 강력한 기회를 제공하고 있다. 여러 특화된 툴킷이 특정 작업에서 높은 성능을 제공하지만, 기존 도구를 사용하는 데에는 여전히 상당한 기술적 전문성이 요구된다. 다양한 환경에서 비전문가에게 충분한 추적 성능을 제공하는 접근 가능한 통합 솔루션은 아직 부재한 상태다. TrackStudio는 이러한 격차를 해소하기 위해, 기존 오픈소스 도구들을 즉시 사용 가능한 단일의 모듈식 GUI 기반 파이프라인으로 통합하여 개발되었다. 이 툴킷은 프로그래밍 기술 없이도 비디오 녹화 전처리, 녹화 동기화, 자동 2D 및 3D 포즈 추정, 시각화를 제공한다. 우리는 비디오 촬영, 카메라 캘리브레이션, 비디오 동기화, 실험 설정에 대한 실용적인 조언을 담은 사용자 가이드와 함께 일반적인 문제점 및 이를 방지하는 방법에 대한 문서도 제공한다. 툴킷의 성능을 검증하기 위해, 저가형 웹캠 또는 고해상도 카메라를 사용하여 신체 위치, 조명, 공간, 장애물 등 까다로운 조건을 포함한 세 가지 환경에서 성능을 테스트하였다. 76명의 참가자를 대상으로 한 실험에서 평균 프레임 간 상관관계는 0.98을 초과하였고, 평균 삼각측량 오차는 낮게 유지되어(손 추적의 경우 13.6mm 미만) 안정적이고 일관된 추적 성능을 입증하였다. 또한 동일한 파이프라인이 손 추적을 넘어 다른 신체 및 얼굴 부위로 확장 가능함을 보여준다. TrackStudio는 전문적 전문성 없이도 신뢰할 수 있는 성능이 필요한 연구자나 일반인을 위해 마커리스 추적에 접근할 수 있는 실용적인 경로를 제공한다.
DUSt3R, MASt3R, VGGT, $π^3$, Depth Anything 3와 같은 최근 3D 기반 모델들은 핀홀 이미지에서 강력한 피드포워드 깊이 및 포즈 추정을 제공하지만, 피시아이 카메라 기하학에서는 성능이 급격히 저하된다. 본 연구는 이러한 실패가 부분적으로 사전학습된 3D 기반 모델의 위치 인코딩에 내재된 핀홀 카메라 바이어스에 기인함을 보이고, 경량 카메라 적응 접근법인 RayTun3R을 제안한다. RayTun3R은 사전학습된 네트워크를 고정한 채 토큰 위치 및 카메라 기하학과 연관된 경량 구성요소만 적응시킨다. 이 방법은 절대적 위치 인코딩과 회전 위치 인코딩에 대한 매개변수 효율적인 잔차 보정을 학습하며, 잔여 핀홀 가정을 제거하는 매개변수 없는 토큰화 및 예측 그리드 좌표 보정을 함께 수행한다. 결과로 얻는 어댑터는 학습 가능한 매개변수가 10,752개에 불과하며, 기하학적 손실을 이용해 짧은 시간 구간에서 학습할 수 있다. 적응 후 RayTun3R은 추가적인 런타임 비용 없이 시퀀스의 나머지 프레임에 효과적으로 전이된다. 시야각이 $110^\circ$에서 $200^\circ$까지 다양한 피시아이 데이터셋에서, 본 어댑터는 적응되지 않은 모델 대비 회전 오차를 $2$~$12$배 감소시키고, 학습 가능한 매개변수를 약 $14$배 적게 사용하면서도 LoRA를 능가하며, 다중 뷰 추론 비용을 회피하면서도 적응 없는 기준선보다 포즈 성능을 향상시키고, 깊이 정확도에서도 경쟁력을 유지한다.