15
SMART RESOURCE-AWARE MULTI-SENSOR NETWORK INTERREG IV RESEARCH PROJECT Klagenfurt, September 2, 2011 MASSIMILIANO VALOTTO PAOLO OMERO SABRINA LONDERO 1 Autonomous complex event detection in scenarios with limited infrastructure valo’[email protected] [email protected] [email protected] h’p://www.infofactory.it

SRS-NET Smart Resource Aware Multi Sensor Network

Embed Size (px)

Citation preview

Page 1: SRS-NET Smart Resource Aware Multi Sensor Network

SMART RESOURCE-AWARE MULTI-SENSOR NETWORK INTERREG IV RESEARCH PROJECT

Klagenfurt, September 2, 2011 MASSIMILIANO VALOTTO PAOLO OMERO SABRINA LONDERO

1  

Autonomous complex event detection in scenarios with limited infrastructure  

valo'[email protected]  -­‐  [email protected]  -­‐  [email protected]  h'p://www.infofactory.it  

Page 2: SRS-NET Smart Resource Aware Multi Sensor Network

%

2  

Designing a smart resource-aware MULTISENSOR NETWORK capable of autonomously DETECTING and LOCALIZING various EVENTS such as screams, animal noise, tracks of PERSONS and more COMPLEX HUMAN BEHAVIOURS."      

MAIN GOAL : SMART MULTISENSOR NETWORK

Page 3: SRS-NET Smart Resource Aware Multi Sensor Network

RESEARCH AREAS

%

3  

1. NETWORK RECONFIGURATION

Due  to  limited  resources,  the  sensors  network  should  be  able  to  reconfigure  itself  in  order  to  limit  consumes  (for  example  switching  off  cameras  when  nothing  happens  in  that  area).  

2. AUDIO/VIDEO ANALISYS Video  frames  and  audio  signals  are  analyzed  in  order  to  recognize  objects  and  sounds.  We  can  idenKfy  for  example  the  type,  speed,  direc2on  and  the  coordinates  of  a  moving  object.  It  is  possible  to  recognize  different  classes  of  objects  such  as  humans,  cars,  dogs  and  cows.  

3. COMPLEX EVENT DETECTION

Seman2c  analysis  is  performed  over  data  extracted  during  audio  and  video  analysis,  in  order  to  detect  complex  events,  such  as  for  example    <people  shoo2ng  to  deers>    <person  walking  in  a  restricted  area>  <dog  figh2ng  with  person>    For  this  purpose  we  use  an  ontological  model  and  a  rules  engine.  

4. MULTIMEDIA DB, RETRIEVAL & ANALYSIS The  MulKMedia  DB  is  devoted  to  archive  the  video  and  audio  files  received  from  sensors.  Furthermore  the  system  is  consKtuted  by  an  advanced  access  &  retrieval  &  knowledge-­‐discovery  layer  

Page 4: SRS-NET Smart Resource Aware Multi Sensor Network

NETWORK

ACQUISITION

ANALYSIS

COLLECTING

DATA MINING

SOUND DETECTION OBJECT RECOGNITION LOCALIZATION

SEMANTIC ANALISYS

MULTIMEDIA & EVENTS ARCHIVE COMPLEX EVENT DETECTED

SOLAR POWERED AUTO RECONFIGURABLE

VIDEO AUDIO PICTURES

Page 5: SRS-NET Smart Resource Aware Multi Sensor Network

1. NETWORK RECONFIGURATION

 Change  power  mode  of  nodes  and  components        Find  op2mal  resource  alloca2on  in  the  network      Move  cameras  in  order  to  follow  the  scene  of  ac2on  and  switch  on  a  camera  when  something  is  expected  to  happen  in  a  specific  area  

Operate the network at highest possible performance while minimizing resource usage."

 Dynamically  adapt  network  structure  and  node  configura2on  according  to  current  applica2on  requirements    LOW  ACTIVITY  à  exchange  only  status  informa2on,  power  down  as  many  sensors  as  possible    HIGH  ACTIVITY  à  exchange  control  and  data  messages,  ac2vate  as  much  sensors  as  needed  

Page 6: SRS-NET Smart Resource Aware Multi Sensor Network

2. AUDIO & VIDEO ANALYSIS

Classifica2on  of  audio  sources.    Iden2fy  specific  sound  paRerns  based  on  characteris2c  features    Examples:  barking  dogs,  shou2ng  humans    

3D Localization, recognition and classification of audio sources. "

Localiza2on  of    sound  sources  with  2me  difference  of  arrival  (TDOA)    

waves  hit  the  microphones  at  different  2me  instances  TDOA  is  related  to  the  line  of  origin  of  the  sound  wave    

Page 7: SRS-NET Smart Resource Aware Multi Sensor Network

2. AUDIO & VIDEO ANALYSIS

Detect  simple  paRerns  of  ac2vity  on  a  ground  map.    Cover  the  paRerns  with  conic  sec2ons  represen2ng  the  observed  zone  for  each  video  sensor    

Analysis and PTZ-Cameras re-configuration. "

SOLUTION:    Project  real  world  on  camera-­‐based  reference  system      The  new  configura2on  op2mally  covers  the  area  wrt.  the  ac2vi2es  occurring  in  it.    

Page 8: SRS-NET Smart Resource Aware Multi Sensor Network

3. COMPLEX EVENT DETECTION

Define  simple  and  complex  events  by  means  of  a  consistent  ontology    Describe  the  events’  context,  ie.,  spa2al,  temporal,  object  and  event  rela2onships      Apply  reasoning  mechanisms  to  iden2fy  complex  events  from  low  level  features    

Detect simple and complex events by means of a consistent ontology. "

Page 9: SRS-NET Smart Resource Aware Multi Sensor Network

4. MULTIMEDIA DATA BASE, RETRIEVAL & ANALYSIS

Store  mul2media  data,  low  level  features,  simple  and  complex  events  in  a  mul2media  database      Provide  user  interface  for  operators  –  High-­‐level  view  of  “what  is  going  on“      Formulate  complex  queries  (e.g.,all  events  in  a  certain  area,  the  areas  most  frequented  by  bears,  the  sensors  less  ac2ve,  …)    

Collect multimedia data from each sensor, save events, and perform advanced analysis."

Find  paRerns  in  data    Recurring  events  (e,g.  Visitors  are  used  to  stop  in  a  specific  area)  Find  rela2ons  between  events  (event  “a  deer  is  detected  in  the  morning  in  AREA  1”  is  ocen  followed  by  “the  deer  is  detected  in  AREA  2  in  the  acernoon”)à  path  discovery    

Alert  an  operator  Alert  an  operator  using  mobile  devices.    

Provide  a  mobile  interface  to  access    the  event  descrip2on  and  the  audio/video  data    

Page 10: SRS-NET Smart Resource Aware Multi Sensor Network

The person (hunter) is detected by a camera"

"A shot is detected by a microphones array in the same area"

A camera recognizes a deer"

The position of the hunter is computed"

The network is reconfigured to look at the hunter position"

The system alerts an operator and sends the event description “a hunter shot a deer” and the audio/video data"AN

EXAM

PLE O

F THE

EVEN

T DET

ECTIO

N PRO

CESS

Page 11: SRS-NET Smart Resource Aware Multi Sensor Network

POWER SEARCH.

11  

The user interface allows users to perform powerful retrieval operations over the collected data and advanced statistical analysis to get knowledge from the archive. The basic access metaphor used for querying the archive is a what/where/when three dimensional space.

Page 12: SRS-NET Smart Resource Aware Multi Sensor Network

EVENTS.

12  

The search results are visualized and can be navigated following an event/place/network three dimensional approach. The events view shows the list of events resulted from the search. For each event we can see the date, the involved subjects, the action and, if defined, the zone where it happened. We can also see a map showing the exact position of the event and any related multimedia content (videos, images or audio).

Page 13: SRS-NET Smart Resource Aware Multi Sensor Network

DATA MINING.

13  

The application offers to the user also some advanced statistical analysis, useful to get knowledge from the archive. Some examples regard the distribution of events of different types over time/in specific periods or the trend of the activity of sensors.

Page 14: SRS-NET Smart Resource Aware Multi Sensor Network

MOBILE ACCESS.

14  

Page 15: SRS-NET Smart Resource Aware Multi Sensor Network

PROJECT PARTNERS

15  

h'p://www.uni-­‐klu.ac.at  

h'p://www.eye-­‐tech.it/  

h'p://www.lakeside-­‐labs.com/  

h'p://www.infofactory.it/