Browsing by Author "Tseng, Chin-Yuan"

Now showing 1 - 4 of 4

Detection of Moving Features Using IMU-Camera without Knowing Both the Initial Conditions and Gravity Direction
(2013-01-01)
Hu, Jwu-Sheng
;
Tseng, Chin-Yuan
;
Chen, Ming-Yuan
;
電機資訊學士班
;
Undergraduate Honors Program of Electrical Engineering and Computer Science
Detecting moving features relative to ground in the images of a moving camera is important for mobile robot localization in practice. This problem is particularly difficult if the initial conditions of the camera are unknown. In this paper, we propose a moving feature detection method by using a calibrated IMU-camera in a dynamic environment. The proposed method is able to separate static and dynamic features without knowing the IMU-camera initial conditions, as well as the gravity direction. In the method, an estimator initialization algorithm is implemented first to estimate the moving velocity and 3D positions of the feature points, and the gravity direction. Then, a recursive moving object detection algorithm is designed to classify the static and dynamic features based on feature re-projection. The simulation results show that the moving features can be grouped effectively, and the remaining static feature points can be used for camera pose and velocity estimation in a real scale to the ground.
IMU-Assisted Monocular Visual Odometry Including the Human Walking Model for Wearable Applications
(2013)
Hu, Jwu-Sheng
;
Tseng, Chin-Yuan
;
Chen, Ming-Yuan
;
Sun, Kuan-Chun
;
電控工程研究所
In this paper, we present a novel approach for monocular visual odometry combined with an inertial measurement unit to estimate the human walking trajectory. The scale factor is computed from the walking speed estimation derived by the kinematic model of human walking. This model not only keeps the important characteristic in biped rolling-foot model, but also makes the speed estimation feasible using human body acceleration. A wearable and calibrated IMU-Camera device is made to mount on the center of human\'s waist for experimental verification. With the human walking speed and gyroscope measurement, an Unscented Kalman filter is implemented to estimate the current scale factor and refine the attitude of the camera. Experiments in an indoor environment are conducted to show the effectiveness of the proposed method.
Locating Anchors in WSN Using a Wearable IMU-Camera
(2014)
Hu, Jwu-Sheng
;
Huang, Yu-Lun
;
Tseng, Chin-Yuan
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Localization in a wireless sensor network (WSN) becomes important for many modern applications, like landslide detection, precision agriculture, health care, and so forth. The more precise the position of an anchor node is, the more accurate the localization of a sensor node can be measured. Since the Global Positioning System (GPS) device cannot work properly indoor, some existing localization methods configure anchor nodes in a manual fashion. However, neither applying GPS modules nor manually configuring anchor nodes is suitable for a WSN and especially artificial errors ofmanual configuration may be propagated and affect the results of localization. In this paper, we propose an alternative method to estimate anchor node locations in an indoor environment. We collect the Received Signal Strength Indicator (RSSI) data from the anchor node when human is walking around them. Meanwhile, we use a wearable IMU-camera device to assist the moving trajectory estimation. We implement a monocular Visual Odometry with a human walking model to estimate moving trajectories. An Unscented Kalman Filter (UKF) is used to estimate the anchor node location by fusing the RSSI data and moving trajectory. The experiment results show that the proposed method has lower estimation error when locating anchors.
Unscented Blind Image De-blurring Using Camera with Inertial Measurement Unit
(2012)
Tseng, Chin-Yuan
;
Chen, Jian-An
;
Hu, Jwu-Sheng
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Image blur resulting from camera motion is an annoying factor for robotic vision, especially for high-speed applications. This work proposes a sensor fusion model for blind image de-blurring using inertial measurement unit. The model attempts to observe the camera motion, estimate the point spread function and de-convolute the image simultaneously. To solve the problem, an iterative estimation procedure using Maximum A-Posteriori Expectation-Maximization (MAP-EM) algorithms and Unscented Kalman Filter are proposed. Simulation results show the feasibility of the proposed formulation to blindly de-blurring the image under camera motion.