Browsing by Author "Wang, Chieh-Chih"

Now showing 1 - 7 of 7

A two-stage sampling for robust feature matching
(2018-08-01)
Chou, Chih-Chung
;
Seo, Young Woo
;
Wang, Chieh-Chih
;
電機工程學系
;
Department of Electrical and Computer Engineering
For any visual feature-based SLAM (simultaneous localization and mapping) solutions, to estimate the relative camera motion between two images, it is necessary to find correct correspondence between features extracted from those images. Given a set of feature correspondents, one can use a n-point algorithm with robust estimation method, to produce the best estimate to the relative camera pose. The accuracy of a motion estimate is heavily dependent on the accuracy of the feature correspondence. Such a dependency is even more significant when features are extracted from the images of the scenes with drastic changes in viewpoints and illuminations and presence of occlusions. To make a feature matching robust to such challenging scenes, we propose a new feature matching method that incrementally chooses a five pairs of matched features for a full DoF (degree of freedom) camera motion estimation. In particular, at the first stage, we use our 2-point algorithm to estimate a camera motion and, at the second stage, use this estimated motion to choose three more matched features. In addition, we use, instead of the epipolar constraint, a planar constraint for more accurate outlier rejection. With this set of five matching features, we estimate a full DoF camera motion with scale ambiguity. Through the experiments with three, real-world data sets, our method demonstrates its effectiveness and robustness by successfully matching features (1) from the images of a night market where presence of frequent occlusions and varying illuminations, (2) from the images of a night market taken by a handheld camera and by the Google street view, and (3) from the images of a same location taken daytime and nighttime.
Multiple target tracking in occlusion area with interacting object models in urban environments
(2018-05-01)
Chen, Jiun-Fu
;
Wang, Chieh-Chih
;
Chou, Cheng-Fu
;
資訊工程學系
;
Department of Computer Science
Multiple target tracking in crowded urban environments is a daunting task. High crowdedness complicates motion modeling, and occlusion makes tracking difficult as well. Based on the variable-structure multiple -model (VSMM) estimation framework, this paper extends an interacting object tracking (IOT) scheme with occlusion detection and a virtual measurement model for occluded areas. IOT is composed of a scene interaction model and a neighboring object interaction model. The scene interaction model considers the long-term interactions of a moving object and surroundings, and the neighboring object interaction model considers three short-term interactions. With these interacting object models, the motion feature of a moving object can be represented with the weight of each model. A virtual measurement model is proposed to exploit the motion feature with the IOT scheme under occlusion. The proposed approach was validated using a stationary 2D LIDAR. To verify in occlusion, a 3D LIDAR based benchmark system was developed to extract occluded moving segments. The ample experimental results show that the proposed IOT scheme tracks over 57% of occluded moving objects in an urban intersection. (C) 2018 Elsevier B.V. All rights reserved.
Orientation constraints for Wi-Fi SLAM using signal strength gradients
(1970-01-01)
Yen, Hsiao-Chieh
;
Wang, Chieh-Chih
;
Chou, Cheng-Fu
;
電機工程學系
;
Department of Electrical and Computer Engineering
We propose the signal strength gradient (SSG) orientation constraints for simultaneous localization and mapping (SLAM) using Wi-Fi received signal strength (RSS) measurements. We show that under certain circumstances, the relative orientation between nearby trajectory segments can be recovered from the cosine similarity between their SSGs. We then show how to obtain trajectory segments and self-consistent SSGs by jointly segmenting Wi-Fi measurements and odometry. Because SSG orientation constraints inevitably contain outliers, we also evaluate the effectiveness of robust SLAM backends on the proposed constraints. Experiments show that Wi-Fi SLAM using the proposed method can correctly estimate orientations given topologically incorrect initialization on trajectories with little to no overlapping sections.
Orientation constraints for Wi-Fi SLAM using signal strength gradients (vol 56, pg 812, 2020)
(2020-09-01)
Yen, Hsiao-Chieh
;
Wang, Chieh-Chih
;
Chou, Cheng-Fu
;
電機工程學系
;
Department of Electrical and Computer Engineering
Probabilistic Structure from Sound
(2009)
Wang, Chieh-Chih
;
Lin, Chi-Hao
;
Hu, Jwu-Sheng
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Auditory perception is one of the most important functions for robotics applications. Microphone arrays are widely used for auditory perception in which the spatial structure of microphones is usually known. In practice, microphone array calibration can be tedious and other devices or means are required. The structure from sound (SFS) approach addresses the problem of simultaneously localizing a set of microphones and a set of acoustic events that provides a great flexibility to calibrate different setups of microphone arrays. However, the existing method does not take measurement uncertainty into account and does not provide uncertainty estimates of the SFS results. In this paper, we propose a probabilistic structure from sound (PSFS) approach using the unscented transform in which the uncertainties of the PSFS results are also available. In addition, a probabilistic sound source localization approach using the PSFS results is provided to improve sound source localization accuracy. The ample results of simulation and experiments using low-cost, off-the-shelf microphones demonstrate the feasibility and performance of the proposed PSFS approach. (C) Koninklijke Brill NV, Leiden and The Robotics Society of Japan, 2009
Simultaneous Heterogeneous Sensor Localization, Joint Tracking, and Upper Extremity Modeling for Stroke Rehabilitation
(2020-09-01)
Chen, Jiun-Fu
;
Wang, Chieh-Chih
;
Wu, Eric Hsiao-Kuang
;
Chou, Cheng-Fu
;
電機工程學系
In stroke rehabilitation systems and applications, reliability, accuracy, and occlusion should be taken into consideration. Unfortunately, most existing approaches focus primarily on the first two issues. However, during the stroke rehabilitation process, occlusion leads to incorrect judgements even for medical staff. In order to tackle these three important issues simultaneously, we propose a heterogeneous sensor fusion framework composed of an RGB-D camera and a wearable device to consider occlusion and provide robust joint locations for rehabilitation. To fuse multiple sensor measurements when compensating for occlusion, we apply heterogeneous sensor simultaneous localization, tracking, and modeling to estimate the locations of joints and sensors and construct an upper extremity model for occlusion situation. Virtual measurements based on this model are used to estimate the joint's location during occlusion, and a virtual relative orientation technique is applied to relax system limitations regarding orientation. Experimental results using the proposed approach with synthetic data and data collected from ten subjects show a 4.6 cm error on average and about 15 cm error on average during occlusion. This constitutes a more robust approach for stroke patients which takes into account these three important issues.
Simultaneous Localization of Mobile Robot and Multiple Sound Sources Using Microphone Array
(2009)
Hu, Jwu-Sheng
;
Chan, Chen-Yu
;
Wang, Cheng-Kang
;
Wang, Chieh-Chih
;
電控工程研究所
Sound source localization is an important function in robot audition. The existing works perform sound source localization using static microphone arrays. This work proposes a framework that simultaneously localizes the mobile robot and multiple sound sources using a microphone array on the robot. First, an eigenstructure-based generalized cross correlation method for estimating time delays between microphones under multi-source environments is described. A method to compute the far field source directions as well as the speed of sound using the estimated time delays is proposed. In addition, the correctness of the sound speed estimate is utilized to eliminate spurious sources, which greatly enhances the robustness of sound source detection. The arrival angles of the detected sound sources are used as observations in a bearings-only SLAM procedure. As the source signals are not persistent and there is no identification of the signal content, data association is unknown which is solved using FastSLAM. The experimental results demonstrate the effectiveness of the proposed approaches.