|
A Comparison between Stereo and Time-Differential Depth estimation for Vision/GPS/INS |
|
Yu Dam Lee, Hyung Keun Lee(Korea Aerospace University, Korea) |
| This paper proposes a method to obtain absolute coordinates of point clouds on designated objects in consecutive images utilizing Global Positioning System (GPS), an Inertial Navigation System (INS) and a stereo camera. A Real-Time Kinematic (RTK) software is utilized to generate accurate absolute coordinates of the camera. A GPS/INS Kalman filter is used to estimate absolute coordinates and attitude parameters of the camera. Due to insufficient accuracy of low-cost Inertial Measurement Unit (IMU), yaw angle error is estimated independently by matched feature points in two consecutive images. |
|
|
Noise Covariance Adaptation of Doppler Velocity Log Considering Vehicle's Attitude |
|
Jaehyuck Cha(Seoul National University, Korea), Hojin Ju(Automation and Systems Research Institute, Korea), Chan Gook Park(Seoul National University, Korea), Kijeong Yoo, Chanju Park(Agency for Defense Development, Korea) |
| This paper analyzes the error characteristics of Doppler velocity log measurements and proposes a noise covariance adaptation method by using the vehicle's attitude information. Then, the measurement is used to suppress the error of inertial navigation system. The analysis was verified through computer simulations. |
|
|
Pedestrian Dead Reckoning System Using Dual IMU to Consider Heel Strike Impact |
|
Hojin Ju(Automation and Systems Research Institute, Korea), Jae Hong Lee, Chan Gook Park(Seoul National University, Korea) |
| In this paper, we propose pedestrian dead-reckoning (PDR) system using shoe and calf mounting inertial measurement units (IMUs) to consider heel strike impact. Heel strike impact causes a large acceleration and angular velocity that cannot be measured by the accelerometer and the gyroscope. Generally, the designed extended Kalman filter (EKF) does not correctly reflect the actual environment, because conventional algorithms do not take into consideration the non-measurable acceleration. In order to consider heel strike impact, we proposed the PDR system using dual IMU with joint constraint. |
|
|
Vehicle Localization Using Road Marking Image Matching |
|
Yeong-Dong Choi(Konkuk University, Korea) |
| In this paper, we propose a method which matches 2D-NDT map with road marking image for fast and accurate vehicle positioning. This method extracts features from occupancy grid map by Maximally Stable Extremal Region (MSER) detector. The extracted features on the map is modelled to centroid and covariance and these are stored on the map. Road markings also need to be extracted by MSER in a bird's eye view camera image. Vehicle Localization is performed by NDT map matching between the map and road markings in a bird's eye view road image. |
|