90 likes | 213 Vues
This proposal aims to enhance the CCIL LifeLog system through improved visualization techniques. Our current tool lacks the capability to effectively present the extensive data we collect from a ubiquitous sensor system analyzing human behavior. We envision creating a more impactful and immersive experience using real-time 3D graphics. The project includes the integration of sensor data into 3D environments for both desktop and mobile applications, allowing users to visualize and interact with their data meaningfully, facilitating better insights and understanding.
E N D
Proposal for next collaboration on“LifeLog visualization” NEC C&C Innovation Research Lab. Kazuo KUNIEDA Nov. 2, 2009
Motivation • To enhance CCIL LifeLog system. We have a ubiquitous-sensors system for human behavior analysis. We are continuously gathering a lot of data from it. However, we have only a simple tool for visualization. • To achieve more imactful, immersive, realistic, by realtime 3D graphics. Simple viewer Advanced viewer
CCIL environment • LifeLog data • All data are stored on Web server and RDB. • Raw data: XML format on Web Server. • Pre-processed data: Binary format on PostgreSQL. • Fundamental 3D data 3D-CAD data (for 3DSMax)
LifeLog Data in CCIL • People’s flow information on a 1.5 m mesh grid with infrared tag. • Acoustic information on a 3 m grid (Ceiling Microphone) • PC key operation, process information, window information, file access information (open/close timing) • Video footage encompassing shooting the indoor environment (Ceiling Video Camera) • Location detection information of books using radio frequency identification (RFID) • Footprints information on 2.5cm mesh grid in corridor (Floor sensor) partially structured Postgre SQL Web Server
Goal: Two aspects of visualization • 3D CG Perspective office view • for desktop client • Sensor data overlaid on 3D CG office view. • realtime view andreplay view • AR view • For mobile client (iPhone, iTouch, etc..) • Sensor data overlaid on the camera imagery displayed on the client screen. • For detection of client location • Ex. using ubiquitous markers like 2D barcode placed in the room.
Object-Location Sensing • To detect both object-location and orientation • Using AR toolkit and pan-tilt type cameras on the ceiling • AR toolkit provides easy API to calculate the real world location of objects using 2D barcode. This feature is optional, but very attractive.
Target user • CCIL researcher • visiting researcher • temporary guest
NEC member • Kazuo KUNIEDA • Hideki KAWAI • Takao Shime • Yuki KAMIYA