The videos in this page show some of the research performed by the LOPSI group:
IMUE-CSIC: Inertial System for Gait Analysis
The IMUE-CSIC system allows easy, quick and accurate gait analysis through the use of inertial sensors placed on the participant’s feet. This solution includes the IMUE-CSIC_GaitAnalysis_Lab software, which provides a detailed report on gait characteristics and parameters.
The VIVIFIL mobile app (registered trademark) provides an 8-week program of personalized physical exercise with audios, animations and video tutorials, for 4 skill levels according to the participant’s abilities. This exercise program has been designed by Dr. Rafael García Molina from Albacete Hospital. The VIVIFIL app also allows direct communication between participants and supervisors (health personnel) to encourage adherence to the program and increase the improvement of the participants’ health
Indoor positioning using PDR, UWB and RFID measurements fused with a particle filter
Using a particle filter we are able to fuse the information of an IMU with the RSS measurements from 34 RFID tags and the TDOA and AOA measurements from 6 Ubisense UWB sensors. The person carries an IMU in the foot, an RFID reader in the waist and an UWB tag in the chest. All the measurements (except the UWB) were recorded with an android device.
Light-Matching: improving PDR using light detections from unmodified lamps in a building
Light intensity measurements from unmodified lamps in our building using the illumination sensor in a smartphone are used to correct a foot-mounted IMU PDR position estimate with a particle filter method. The phone’s GPS measurements are also used outdoors, however, no map-matching, WiFi, RFID or magnetometer are used in the trajectory of the video.
Indoor/outdoor pedestrian localization
Shows the results of an Indoor/Outdoor Pedestrian Localization experiment using a Particle Filtering fusion approach integrating: Map-matching, Inertial foot-mounted Dead-Reckoning, GPS/Glonass outdoors positioning and a S3 Samsung Smartphone.
Real Test time: 20 minutes. Test length: 1.3 km. Positioning Accuracy below 1 meter.Video accelerated 4x
IMU signals at foot stance
In pedestrian navigation (PDR), it’s assumed that the velocity of the foot is zero during the stance of the foot on the ground. Here we can visualize the actual experimental accelerometer, gyroscope and magnetometer signals from a foot-mounted IMU during the stance period.
Signals are sampled at 100 Hz, and the video frames at 25 Hz. The colors used in the plots (Red, Green, Blue) represent the IMU axes, respectively: X (forward), Y (lateral/left) and Z axis (up).
Improved Heuristic Drift Elimination (iHDE) for IMU-based Indoor Pedestrian Navigation
Dead-reckoning position estimation of a person walking in buildings. The system uses 3-accelerometes and 3 gyroscopes (IMU) installed on the person’s foot. Heading is corrected using the building’s dominant directions, so no magnetic sensors are needed (avoiding the use of the typically perturbated local Earth magnic field). Positioning error is lower than 0.3% of Total Travelled Distance (TTD), e.g. the trajectory of this video accumulates 1.5 metres positioning error in a 600-meter-long path. ( Video speed is 3x). More information: download publication pdf.
Indoor Pedestrian Localization with IMU and RFID technology
Localization of a person walking in an indoor building using a foot-mounted IMU, with assistance of an RFID system deployed in the building (the RFID reader is carried by the person in his belt). Download publication PDF.
Indoor Pedestrian localization with RFID technology
Indoor localization of a person using RF signals from RFID tags placed at known positions in our building. The current position is displayed in a PDA. Demo video presented as final report of the DEDALO project.
Ultrasonic indoor localization
A video of the prototype of the 3D-Locus ultrasonic positioning system, showing the beacon network, the mobile node, and the user interface, demonstrating the subcentimeter positioning in three dimensions using ultrasound trilateration.
Fruit Picking by a robot
The fruit is located using a laser tracker and image processing software, and the estimated position is used by a robotic arm to collect the piece of fruit .
Ultrasonic Localization of Findings in Archaeology
The 3D-Locus ultrasonic positioning system deployed at the Atapuerca archaeological site, as an assistance aid for determination of the exact coordinates of fossils and other finds. Shows the traditional positioning methods used by the archaelogical workers and how the process is automated with the 3D-Locus