KES: Knowledge Enabled Services for better EO Information Use - - PowerPoint PPT Presentation

▶
kes knowledge enabled services for better eo information
SMART_READER_LITE
LIVE PREVIEW

KES: Knowledge Enabled Services for better EO Information Use - - PowerPoint PPT Presentation

Feb 14, 2023 104 likes •429 views

KES: Knowledge Enabled Services for better EO Information Use Andrea Colapicchioni Advanced Computer Systems Space Division a.colapicchioni@acsys.it The problem During the last decades, the satellite image catalogues have stored huge

slide-1
SLIDE 1

KES: Knowledge Enabled Services for better EO Information Use

Andrea Colapicchioni Advanced Computer Systems Space Division a.colapicchioni@acsys.it

slide-2
SLIDE 2

IGARSS 2004 - Image Information Mining

The problem

During the last decades, the satellite

image catalogues have stored huge quantity of data

State of the art catalogues permit only

to specify location, time of interest, metadata like platform, sensor, acquisition mode…

slide-3
SLIDE 3

IGARSS 2004 - Image Information Mining

The interpretation task

The interpretation of EO images

requires

Fusion of data/information for better

understanding of structures

Aggregation with existing knowledge

specific to the application fields (at higher level)

slide-4
SLIDE 4

IGARSS 2004 - Image Information Mining

A little bit of history..

1996-2000 IIM (Image Information Mining)

(http://isis.dlr.de/mining)

2001: KIM (Knowledge Information Mining)

(http://www.acsys.it:8080/kim)

2002: KES (Knowledge Enabled Services) 2003: KIMV (KIM Validation) 2004: KEO (Knowledge-centred Earth Observation)

1996 2006 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 1996 - 2000 IIM (DLR, ETH) 2001 - 2002 KIM (ACS, DLR, ETH) 2002 - 2004 KES (DLR, ACS) 2004 - 2006 KEO (ACS, DLR, GTD, CNES) 2003 - 2004 KIMV (ACS, DLR)

slide-5
SLIDE 5

IGARSS 2004 - Image Information Mining

KIM: Knowledge Driven Information Mining

Images Primitive Feature Extraction (at pixel level) 010010 101101 110010 010010 101101 110010 Primitive Features:

  • Texture
  • Colour
  • Shape
slide-6
SLIDE 6

IGARSS 2004 - Image Information Mining

More in detail

slide-7
SLIDE 7

IGARSS 2004 - Image Information Mining

KIM Interactive learning

slide-8
SLIDE 8

IGARSS 2004 - Image Information Mining

From data to semantic

Images Primitive Feature Extraction 010010 101101 110010 010010 101101 110010 Primitive Features:

  • Texture
  • Colour
  • Shape

Interactive Learning River Forest Cloud Features DATA SEMANTIC

slide-9
SLIDE 9

IGARSS 2004 - Image Information Mining

From KIM to KES (Knowledge Enabled Services)

Image interpretation is not a simple task.

Each user needs a set of accessory data, as for example GIS layers or texts obtained through Internet.

Yet, the amount of available information

makes searches a demanding and expensive task.

An environment where images are at the

focal point, and where each user can navigate through a taxonomically structured knowledge, could be of extreme value.

slide-10
SLIDE 10

IGARSS 2004 - Image Information Mining

Other semantic

Images

Primitive Feature extraction

Primitive Features Image Features Text GIS Layers

Interactive Learning Interaction Interaction

Text Features GIS Features DATA SEMANTIC

slide-11
SLIDE 11

IGARSS 2004 - Image Information Mining

KES: New interface

slide-12
SLIDE 12

IGARSS 2004 - Image Information Mining

Semantic Grouping

In KIM it is simple to define a feature

by identifying a river through positive and negative examples.

However, the “river" might become

part of a wider concept (e.g.: water). This should be implemented without retraining the system for all possible water types.

slide-13
SLIDE 13

IGARSS 2004 - Image Information Mining

Aggregated Features

In

KES a new kind

  • f

grouping, called “aggregated features”, has been introduced.

In a similar way as in defining positive and

negative examples on the image, it will be possible to define the concept “water” as:

Positive examples:

sea + river + lake + water reservoir

Negative examples:

streets + houses + mountains

slide-14
SLIDE 14

IGARSS 2004 - Image Information Mining

Semantic Grouping

Images

Primitive Feature extraction

Primitive Features Image Features Text GIS Layers

Interaction

Text Features GIS Features DATA SEMANTIC

Semantic Grouping

Aggregated Image Features

Interactive Learning Interaction

Object Descriptor

Descriptive Groupings Descriptive Groupings Descriptive Groupings Descriptive G roupings

slide-15
SLIDE 15

IGARSS 2004 - Image Information Mining

Ontology

Ontology is the specification of a

  • conceptualisation. It can be related to a

system (System Ontology, which can be domain independent and reused for different domains) or to a domain (Domain Ontology, specific for that domain).

slide-16
SLIDE 16

IGARSS 2004 - Image Information Mining

Domain Ontology

Domain ontology is the set of definitions and

concepts pertaining and belonging to a specific domain (and shared by concerned people).

Different domains have generally different

  • ntologies. As an example, a climatology

expert could have a different vision of (and terms to describe) water compared to that of an oceanography expert.

slide-17
SLIDE 17

IGARSS 2004 - Image Information Mining

Taxonomy

Domain 1 Domain n … Sub domain Sub domain Sub domain Aggregated Image Feature Image Feature Image Feature Aggregated Image Feature Image Feature Image Image Image Image Text Feature GIS Feature Text Text GIS Layer GIS Layer

slide-18
SLIDE 18

IGARSS 2004 - Image Information Mining

Descriptive G roupings

Data / Semantic / Ontology

Images Primitive Features Image Features Text GIS Layers

Interaction

Text Features GIS Features

DATA SEMANTIC

Semantic Grouping

Aggregated Image Features

Interaction

Object Descriptor

Descriptive Groupings D e s c r i p t i v e G r

  • u

p i n g s Descriptive Groupings Primitive Feature extraction Interactive Learning

Domains

DOMAIN ONTOLOGY ONTOLOGY

slide-19
SLIDE 19

IGARSS 2004 - Image Information Mining

Implicit and explicit knowledge transfer

The KIM prototype permits to associate weighted

combinations of primitive features to image features. While defining features, the user explicitly transfers knowledge to the system, which is stored and made available to the same or other users.

In addition there is a further type of knowledge

(implicit) that could be discovered: if the user domain

  • f interest is known, by observing the user

interactions with the system during search and browsing, it is possible to infer the data of likely user interest and link it with the pertinence domain.

slide-20
SLIDE 20

IGARSS 2004 - Image Information Mining

Knowledge transfer

Lake Sea Roads

ERS Image Landsat Image Meris Image

Algae Blooming Climate Changes Cities Temperature Census Map

IKONOS Image

Roads Urban Areas Factories Political Facts

slide-21
SLIDE 21

IGARSS 2004 - Image Information Mining

Knowledge Graph

www.a.com MERIS Image www.a.com Spectral (feature) www.a.com Ikonos Image Water GIS Water River River GIS www.b.com Hydrology

slide-22
SLIDE 22

IGARSS 2004 - Image Information Mining

Knowledge Discovery

Exploring this graph means discovering the user's knowledge

slide-23
SLIDE 23

KIMV (KIM Validation)

A practical case study: Automatic cloud classification in MERIS images

slide-24
SLIDE 24

IGARSS 2004 - Image Information Mining

Cloud Cover Characterization

MERIS Level 1 – Reduced Resolution “cloud”

KIMV System

slide-25
SLIDE 25

IGARSS 2004 - Image Information Mining

Cloud cover characterization

Map of Cloud

  • bject
slide-26
SLIDE 26

IGARSS 2004 - Image Information Mining

MAP Object Extraction

Binarized, closed MAP of “cloud” label

slide-27
SLIDE 27

IGARSS 2004 - Image Information Mining

Tiling

(x1, y1)-(x2,y2)… 20 12345 (x1,y1)-(x2,y2)…. 60 12345

Shape Vote I MG

slide-28
SLIDE 28

IGARSS 2004 - Image Information Mining

MASS

slide-29
SLIDE 29

IGARSS 2004 - Image Information Mining

Linux Cluster

Cluster nodes: 10 Dell 1750, dual

Xeon 3 GHz

Database and Application Server:

Dell 6600, 4 Xeon 2.8 GHz

RACK 01

Power Vault (14x146 GB) PowerEdge 1750 (x 5) PowerEdge 2650

6 6 0 F 1 2 3 4 5 6 1 2 3 4 5 6 PowerVault 3 6 G B 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 G B 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 3 6 GB 1 0 K r p m F C 7 1 2 1 6 1 9 2 4 1 3 1 8 3C17203 Super Stack 3 1 2 3 4 5 6 7 8 9 1 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 2 1 2 2 2 3 2 4 1 2 3 4 5 6 7 8 9 1 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 2 1 2 2 2 3 2 4 S t a t u s P a c k e t S t a t u s P a c k e t Packet - Green = Full Duplex Green = 100Mbps Yellow = 10Mbps S t a t u s
  • Yellow = Half Duplex
  • n = enabled, link ok
flashing = disabled 4 2 3 1 8 6 7 5 M
  • d
u l e 1 Module 2 P o w e r / S e l f T e s t Switch 4400 

Giga Switch

1 7 5 P o w e r E d g e X E O N 1 7 5 P o w e r E d g e X E O N 1 7 5 P o w e r E d g e X E O N 1 7 5 P o w e r E d g e X E O N

KVM Switch Rackable Monitor & Keyboard (1U)

slide-30
SLIDE 30

IGARSS 2004 - Image Information Mining

System Context

KI M SYSTEM

Internet Parallel Ingestion Chain Expert user Direct Connection

MASS TOOLBOX

RDBMS

KIM Server

Internet Ingestion chain monitor Direct Connection JDBC WEB SERVICE Ingestion daemon

Workflow Engine

User Client Internet MERIS ARCHIVE FACILITY

MASS PORTAL

slide-31
SLIDE 31

IGARSS 2004 - Image Information Mining

KIM prototype

http://www.acsys.it:8080/kim Andrea Colapicchioni

a.colapicchioni@acsys.it

Visit the ACS stand for a demo