• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.328-335
• /
• 2022

With the advance of robot capabilities and functionalities, construction robots assisting construction workers have been increasingly deployed on construction sites to improve safety, efficiency and productivity. For close-proximity human-robot collaboration in construction sites, robots need to be aware of the context, especially construction worker's behavior, in real-time to avoid collision with workers. To recognize human behavior, most previous studies obtained 3D human poses using a single camera or an RGB-depth (RGB-D) camera. However, single-camera detection has limitations such as occlusions, detection failure, and sensor malfunction, and an RGB-D camera may suffer from interference from lighting conditions and surface material. To address these issues, this study proposes a novel method of 3D human pose estimation by extracting 2D location of each joint from multiple images captured at the same time from different viewpoints, fusing each joint's 2D locations, and estimating the 3D joint location. For higher accuracy, the probabilistic representation is used to extract the 2D location of the joints, considering each joint location extracted from images as a noisy partial observation. Then, this study estimates the 3D human pose by fusing the probabilistic 2D joint locations to maximize the likelihood. The proposed method was evaluated in both simulation and laboratory settings, and the results demonstrated the accuracy of estimation and the feasibility in practice. This study contributes to ensuring human safety in close-proximity human-robot collaboration by providing a novel method of 3D human pose estimation.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.19 no.1
• /
• pp.9-17
• /
• 2024

It is widely recognized that for 3D human pose estimation (HPE), dataset acquisition is expensive and the effectiveness of augmentation techniques of conventional visual recognition tasks is limited. We address these difficulties by presenting a simple but effective method that augments input images in terms of viewpoints when training a 3D human pose estimation (HPE) model. Our intuition is that meaningful variants of the input images for HPE could be obtained by viewing a human instance in the images from an arbitrary viewpoint different from that in the original images. The core idea is to synthesize new images that have self-occlusion and thus are difficult to predict at different viewpoints even with the same pose of the original example. We incorporate this idea into the training procedure of the 3D HPE model as an augmentation stage of the input samples. We show that a strategy for augmenting the synthesized example should be carefully designed in terms of the frequency of performing the augmentation and the selection of viewpoints for synthesizing the samples. To this end, we propose a new metric to measure the prediction difficulty of input images for 3D HPE in terms of the distance between corresponding keypoints on both sides of a human body. Extensive exploration of the space of augmentation probability choices and example selection according to the proposed distance metric leads to a performance gain of up to 6.2% on Human3.6M, the well-known pose estimation dataset.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.40 no.6
• /
• pp.31-40
• /
• 2003

Most face recognition systems are based on 2D images and applied in many applications. However, it is difficult to recognize a face when the pose varies severely. Therefore, head pose estimation is an inevitable procedure to improve recognition rate when a face is not frontal. In this paper, we propose a novel head pose estimation algorithm for 3D face recognition. Given the 3D range image of an unknown face as an input, we automatically extract facial feature points based on the face curvature. We propose an Error Compensated Singular Value Decomposition (EC-SVD) method based on the extracted facial feature points. We obtain the initial rotation angle based on the SVD method, and perform a refinement procedure to compensate for remained errors. The proposed algorithm is performed by exploiting the extracted facial features in the normaized 3D face space. In addition, we propose a 3D nearest neighbor classifier in order to select face candidates for 3D face recognition. From simulation results, we proved the efficiency and validity of the proposed algorithm.

• Journal of Korea Multimedia Society
• /
• v.7 no.12
• /
• pp.1745-1753
• /
• 2004

A method of estimating the pose of a three-dimensional object from a set of two-dimensioal images based on parametric eigenspace method is proposed. A Gaussian blurred edge image is used as an input image instead of the original image itself as has been used previously. The set of input images is compressed using K-L transformation. By comparing the estimation errors for the original, blurred original, edge, and blurred edge images, we show that blurring with the Gaussian function and the use of edge images enhance the data compression ratio and decrease the resulting from smoothing the trajectory in the parametric eigenspace, thereby allowing better pose estimation to be achieved than that obtainable using the original images as it is. The proposed method is shown to have improved efficiency, especially in cases with occlusion, position shift, and illumination variation. The results of the pose angle estimation show that the blurred edge image has the mean absolute errors of the pose angle in the measure of 4.09 degrees less for occlusion and 3.827 degrees less for position shift than that of the original image.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.15 no.5
• /
• pp.31-42
• /
• 2019

With the recent development of deep learning technology, remarkable achievements have been made in many research areas of computer vision. Deep learning has also made dramatic improvement in two-dimensional or three-dimensional human pose estimation based on a single image, and many researchers have been expanding the scope of this problem. The human pose estimation is one of the most important research fields because there are various applications, especially it is a key factor in understanding the behavior, state, and intention of people in image or video analysis. Based on this background, this paper surveys research trends in estimating human poses based on a single image. Because there are various research results for robust and accurate human pose estimation, this paper introduces them in two separated subsections: 2D human pose estimation and 3D human pose estimation. Moreover, this paper summarizes famous data sets used in this field and introduces various studies which utilize human poses to solve their own problem.

• Proceedings of the IEEK Conference
• /
• 2003.07e
• /
• pp.2000-2003
• /
• 2003

This paper presents a three-dimensional (3D) head pose estimation algorithm for robust face recognition. Given a 3D input image, we automatically extract several important 3D facial feature points based on the facial geometry. To estimate 3D head pose accurately, we propose an Error Compensated-SVD (EC-SVD) algorithm. We estimate the initial 3D head pose of an input image using Singular Value Decomposition (SVD) method, and then perform a Pose refinement procedure in the normalized face space to compensate for the error for each axis. Experimental results show that the proposed method is capable of estimating pose accurately, therefore suitable for 3D face recognition.

• Proceedings of the IEEK Conference
• /
• 2007.07a
• /
• pp.211-212
• /
• 2007

Pose-variation factors present a significant problem in 2D face recognition. To solve this problem, there are various approaches for a 3D face acquisition system which was able to generate multi-view images. However, this created another pose estimation problem in terms of normalizing the 3D face data. This paper presents a 3D head pose-normalization method using 2D and 3D interaction. The proposed method uses 2D information with the AAM(Active Appearance Model) and 3D information with a 3D normal vector. In order to verify the performance of the proposed method, we designed an experiment using 2.5D face recognition. Experimental results showed that the proposed method is robust against pose variation.

• Computers and Concrete
• /
• v.33 no.5
• /
• pp.535-544
• /
• 2024

In the rapidly advancing landscape of computer vision (CV) technology, there is a burgeoning interest in its integration with the construction industry. Camera calibration is the process of deriving intrinsic and extrinsic parameters that affect when the coordinates of the 3D real world are projected onto the 2D plane, where the intrinsic parameters are internal factors of the camera, and extrinsic parameters are external factors such as the position and rotation of the camera. Camera pose estimation or extrinsic calibration, which estimates extrinsic parameters, is essential information for CV application at construction since it can be used for indoor navigation of construction robots and field monitoring by restoring depth information. Traditionally, camera pose estimation methods for cameras relied on target objects such as markers or patterns. However, these methods, which are marker- or pattern-based, are often time-consuming due to the requirement of installing a target object for estimation. As a solution to this challenge, this study introduces a novel framework that facilitates camera pose estimation using standardized materials found commonly in construction sites, such as concrete forms. The proposed framework obtains 3D real-world coordinates by referring to construction materials with certain specifications, extracts the 2D coordinates of the corresponding image plane through keypoint detection, and derives the camera's coordinate through the perspective-n-point (PnP) method which derives the extrinsic parameters by matching 3D and 2D coordinate pairs. This framework presents a substantial advancement as it streamlines the extrinsic calibration process, thereby potentially enhancing the efficiency of CV technology application and data collection at construction sites. This approach holds promise for expediting and optimizing various construction-related tasks by automating and simplifying the calibration procedure.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.01a
• /
• pp.277-278
• /
• 2021

최근 손동작 인식에 대한 연구들이 활발하다. 하지만 대부분 Depth 정보를 포함한3D 정보를 필요로 한다. 이는 기존 연구들이 Depth 카메라 없이는 동작하지 않는다는 한계점이 있다는 것을 의미한다. 본 프로젝트는 Depth 카메라를 사용하지 않고 2D 이미지에서 Hand Keypoint Detection을 통해 손동작 인식을 하는 방법론을 제안한다. 학습 데이터 셋으로 Facebook에서 제공하는 InterHand2.6M 데이터셋[1]을 사용한다. 제안 방법은 크게 두 단계로 진행된다. 첫째로, Object Detection으로 Hand Detection을 수행한다. 데이터 셋이 어두운 배경에서 촬영되어 실 사용 환경에서 Detection 성능이 나오지 않는 점을 해결하기 위한 이미지 합성 Augmentation 기법을 제안한다. 둘째로, Keypoint Detection으로 21개의 Hand Keypoint들을 얻는다. 실험을 통해 유의미한 벡터들을 생성한 뒤 Annoy (Approximate nearest neighbors Oh Yeah) Tree를 생성한다. 생성된 Annoy Tree들로 후처리 작업을 거친 뒤 최종 Pose Estimation을 완료한다. Annoy Tree를 사용한 Pose Estimation에서는 NN(Neural Network)을 사용한 것보다 빠르며 동등한 성능을 냈다.

• Journal of Internet Computing and Services
• /
• v.19 no.6
• /
• pp.41-51
• /
• 2018

Recently, in the field of video surveillance, deep learning based learning method is applied to intelligent video surveillance system, and various events such as crime, fire, and abnormal phenomenon can be robustly detected. However, since occlusion occurs due to the loss of 3d information generated by projecting the 3d real-world in 2d image, it is need to consider the occlusion problem in order to accurately detect the object and to estimate the pose. Therefore, in this paper, we detect moving objects by solving the occlusion problem of object detection process by adding depth information to existing RGB information. Then, using the convolution neural network in the detected region, the positions of the 14 keypoints of the human joint region can be predicted. Finally, in order to solve the self-occlusion problem occurring in the pose estimation process, the method for 3d human pose estimation is described by extending the range of estimation to the 3d space using the predicted result of 2d keypoint and the deep neural network. In the future, the result of 2d and 3d pose estimation of this research can be used as easy data for future human behavior recognition and contribute to the development of industrial technology.