INTRODUCTORY REMOTE SENSING€¦ · • Overview of basics of remote sensing: • Electromagnetic spectrum, spectral responses of objects (e.g., trees, houses, crops) • Platforms/Sensors:

INTRODUCTORY REMOTE SENSING

Landsat 7 15 m image highlighting the geology of Oman

http://www.satimagingcorp.com/gallery/landsat-geology-lg.htmlhttp://www.satimagingcorp.com/gallery/landsat-geology-lg.html

ASTER 15 m SWIR image, Geology, Yemen

http://www.satimagingcorp.com/satellite-sensors/aster.htmlhttp://www.satimagingcorp.com/satellite-sensors/aster.html

ASTER 15 m SWIR image, Mexicali

ASTER 15 m image, Las Vegas

MODIS 500 m hyperspectral image, Alaska

https://airbornescience.nasa.gov/instrument/MAShttps://airbornescience.nasa.gov/instrument/MAS

GOALS OF THE COURSE

• Overview of basics of remote sensing: • Electromagnetic spectrum, spectral responses of objects

(e.g., trees, houses, crops) • Platforms/Sensors: where / how they capture energy • How this energy gets converted into images (A2D)• Development of technical / theoretical skills for reading,

enhancing and processing remotely sensed data (DIP)• Use of remotely sensed imagery in environmental change

detection, urbanization, desertification, deforestation, forest fire monitoring, natural hazards monitoring, etc.

WHAT IS REMOTE SENSING?

• Remote sensing is the art and science of acquiring information about the Earth's surface without actually being in contact with it. This is done by sensing and recording reflected or emitted energy and processing, analyzing, and applying that information.

WHAT IS REMOTE SENSING?

• Platforms (C/B/D/A/S)• Data types

• Aerial photographs (analog / digital)• Satellite images (digital)

• Science: Physics• Scattering, etc.• RADAR

• Art: Subjective interpretation

SPECTRAL SIGNATURES AND SENSORS

http://www.tankonyvtar.hu/en/tartalom/tamop425/0032_terinformatika/ch05.htmlhttp://www.tankonyvtar.hu/en/tartalom/tamop425/0032_terinformatika/ch05.htmlhttp://www.tankonyvtar.hu/en/tartalom/tamop425/0032_terinformatika/ch05.htmlhttp://www.tankonyvtar.hu/en/tartalom/tamop425/0032_terinformatika/ch05.html

CHARACTERISTICS OF REMOTELY SENSED IMAGERY

• Data resolutions• Spatial (pixels)• Spectral (bands)• Temporal (orbits)• Radiometric (DNs)

• Data sources: Visible light, near (reflected) to far (heat) IR, microwaves (radar)

• and much more.

False colour

AERIAL PHOTOGRAPHY

• Stereo-effect: pairs of images that are displaced produce a 3-D effect

• Allows for measuring elevation

INFRARED IMAGERY

Airborne thermal IR (25 cm resolution) of homes highlighting energy loss

http://www.imagingnotes.com/go/article_free.php?mp_id=181http://www.imagingnotes.com/go/article_free.php?mp_id=181

SATELLITE IMAGERY: OCEAN MONITORING

Eddies off of Haida Gwaii

SeaWiFS (1 km)June 13, 2002.

http://earthobservatory.nasa.gov/IOTD/view.php?id=2536http://earthobservatory.nasa.gov/IOTD/view.php?id=2536

DROUGHT MONITORING

JULY 2001 JULY 2002

LANDSAT THEMATIC MAPPER (30 m) (SOURCE: CCRS 2002)

(Healthy vegetation is bright red)

MODIS imagery (200 m) showing Myanmar before and after being hit by a cyclone.

DISASTER MONITORING

http://movingimages.wordpress.com/2008/05/

LIGHT DETECTION AND RANGING

Forest canopy (1st return)

Using many rapid small bursts of laser light, a record of the terrain is produced from

the reflections obtained from multiple sources.

Ground surface (last return)

LIDAR

Archaeology

Oceanography

Forest inventory

RADAR – MONITORING SEA ICE

(30 m)

RADAR -- INTERFEROMETRY

THE ‘SYSTEM’

• A: Source of EMR• B: Electromagnetic radiation (EMR)• C: Object of interest• D: Sensor• E: Transmission to receiver• F: Data products (DIP)• G: Results of analyses

A

DIGITAL IMAGE PROCESSING

1. Pre-processing: Radiometric Correction and Georeferencing

2. Classification:1. Unsupervised Classification2. Supervised Classification3. Object-based Classification

3. Change Detection & Data Integration (GIS)

http://www.clarklabs.org/products/Object-Oriented-Image-Segmentation-and-Classification.cfmhttp://www.clarklabs.org/products/Object-Oriented-Image-Segmentation-and-Classification.cfm

DIGITAL IMAGE PROCESSING: CLASSIFICATION

• A wide variety of methods can be used, each of which may produce a different result. No one method is necessarily the ‘best’.

• Accuracy assessment is a necessary component in any classification.• Methods can produce a soft (fuzzy classes) or a hard classification.

Named classes

Training sites

Classified image

Pixels DN-based classesAssign

names to classes

Image segmented into blocks

Objects identified

Blocks assigned to

classes

• Non-uniform• Varies according to

• Wavelength• Time

• Active sensors are less variable

• Uniform emissions

EMR: SOURCE

IDEAL Reality

• Interfering• Varies according to

• Wavelength• Windows

• Atmospheric conditions

• Time and place• Source’s position

• Blue sky• Red sunsets

• Non-interfering

EMR: ATMOSPHERE

IDEAL REALITY

• Common responses from dissimilar objects

• Responses vary even for a known target• Time (e.g., seasons)• Angle of source-target-

sensor• Much remains to be

documented

• Unique• Known

OBJECT REFLECTANCE

IDEAL REALITY

• Sensors have variable sensitivities (and the sensitivity can vary the need for frequent calibration)

• Spatial and spectral resolution varies

• Highly and equally sensitive to all wavelengths

• Consistent spatial and spectral resolutions

SENSORS

IDEAL REALITY

• Processing is not always timely

• Interpretation is fraught with complications

• Results are software- and user-dependent

• Real-time processing and consistent interpretation

• Software is easy to use• Data is easy to obtain

DIGITAL IMAGE PROCESSING

IDEAL REALITY

• Limited or unrealistic impressions of RS

• Lacking in domain-specific knowledge

• Transformation of data to knowledge is complex

• In-depth knowledge of remote sensing and GIS

• Domain-specific knowledge (e.g., forest ecology, oceanography)

USERS

IDEAL REALITY

FINAL EXAM

• Three components:• Definitions (point form okay, use diagrams if helpful)• Short answers (point form okay, use diagrams if helpful)• Essay questions (proper grammar) (50%)

• Exam will cover:• Lectures• Labs• Chapters from text• Entire term

WHAT NEXT?

• Other Geospatial courses in the Department• Advanced GIS: Geob 370• Cartography: Geob 372• Statistics: Geog 374• Advanced Cartography: Geob 472• GIScience in Research: Geob 479

• BCIT• Graduate School• Employment

ALL THE BEST!

Final exam: APR 9 08:30 AM in MCLD 292

http://imgs.xkcd.com/comics/correlation.png

http://maps.ubc.ca/PROD/index_detail.php?show=y,n,n,n,n,y&bldg2Search=n&locat1=312&locat2=#showMapCampushttp://maps.ubc.ca/PROD/index_detail.php?show=y,n,n,n,n,y&bldg2Search=n&locat1=312&locat2=#showMapCampus

Introductory Remote SensingSlide Number 2Slide Number 3Slide Number 4Slide Number 5Slide Number 6Goals of the course What is remote sensing?What is remote sensing?spectral signatures and sensorsCharacteristics of remotely sensed ImageryAerial photographyInfrared ImagerySatellite imagery: Ocean monitoringDrought monitoringDisaster monitoringLight Detection and Ranging LIDARRADAR – Monitoring Sea IceRADAR -- InterferometryThe ‘system’Digital Image ProcessingDigital Image Processing: ClassificationEMR: SourceEMR: AtmosphereObject ReflectanceSensorsDIGITAL IMAGE PROCESSINGUsersSlide Number 30Final examWhat Next?All the best!Slide Number 34