Introduction to Remote Sensing. Electromagnetic Energy. Data From Wave Phenomena. Electromagnetic Radiation (EMR) Electromagnetic Energy

Transcription

1 A Basic Introduction to Remote Sensing (RS) ~~~~~~~~~~ Rev. Ronald J. Wasowski, C.S.C. Associate Professor of Environmental Science University of Portland Portland, Oregon 1 September 2015 Introduction to Remote Sensing The Sabins definition Acquire: Make a permanent record Process: Convert raw data into images Interpret: Convert images into information The Wasowski definition Acquire: Make a permanent record Process: Convert raw data into information Vegetation indices Multispectral classification Interpret: Convert information into action Management decisions, public policy Data From Wave Phenomena Interaction between matter & energy Properties of the matter Electron energy level [shell] characteristics Surface roughness Emissivity Properties of the energy EMR: Wavelength ( color ) Synchronized transverse electric & magnetic oscillations Transmission requires no medium Acoustic: Wavelength ( pitch ) Longitudinal oscillations Transmission requires some medium Electromagnetic Energy Some basic relationships EM propagation: λ. ν = c, where: λ = wavelength (variable) ν = frequency (constant) c = speed of EMR (variable) Temperature Celsius: 0 defined relative to water freezing Called degrees Celsius : Actual temperatures Called Celsius degrees : Temperature differences Kelvins: 0 defined as absolute zero Called kelvins Electromagnetic Radiation (EMR) All energy that moves at light-speed in vacuum ( m. sec -1 ) in a transverse harmonic wave pattern with synchronized perpendicular E & M field vectors Electromagnetic Energy Interaction processes Incident (incoming) radiation Rejection Coherent: Reflection Smooth surface Disorganized: Scattering Rough surface Transmission Coherent: Transparent Clear image Disorganized: Translucent Diffuse image Absorption Continuous: High-density materials Discrete: Low-density materials Emitted (outgoing) radiation Blackbody radiation Spectral lines

2 Electromagnetic Spectrum Remote sensing wavelength regions and bands Band name Wavelengths Notes Gamma rays < 0.03 nm X-rays 0.03 to 30 nm Ultraviolet (UV) 0.03 to 0.4 µm Photographic UV 0.3 to 0.4 µm Film Visible 0.4 to 0.7 µm Sunlight Infrared (IR) 0.7 to 100 µm Reflected (RIR) 0.7 to 3.0 µm Sunlight Thermal (TIR) 3.0 to 14.0 µm»?!?!? Radio Microwave 0.1 to 100 cm Passive Radar 0.1 to 100 cm Active Radio > 100 cm Passive Electromagnetic Spectrum Insolation (Incident solar radiation) Earth s reflected energy peak: ~ 0.5 µm Visible green light Caused by interactions between Earth materials & EMR Emission (Radiated heat energy) Earth s radiant energy peak: ~ 9.7 µm Thermal infrared Expression of blackbody radiation Absorbs 100% of all incident EMR of all wavelengths Emits 100% of absorbed energy based on Kelvin temperature Affected by emissivity Percent efficiency Resolution of Imaging Systems Spatial resolution AVHRR: 1,000 meters per pixel Landsat MSS: 80 meters per pixel Landsat TM: 30 meters per pixel Ikonos: 4 meters per pixel Pléiades: 0.7 meters per pixel Spectral resolution 1 through 255 spectral bands Temporal resolution Landsat MSS & TM: 16 days IRS LISS: 24 days Radiometric resolution 6 bits per pixel: 2 6 = Brightness levels from 0 to 63 7 bits per pixel: 2 7 = Brightness levels from 0 to bits per pixel: 2 8 = Brightness levels from 0 to bits per pixel: 2 12 = Brightness levels from 0 to 4095 Spectral Reflectance Curves R G B NIR MIR The Structure of Digital Images The target & the representation revisited Ground resolution cell The smallest area that can be seen on the ground Picture elements (pixels) The smallest area that can be represented on the image Numerical Values of Digital Images Digital number (DN) Integer values: No decimal point Mathematical sign Unsigned: All values 0 Signed: Negative & positive values Common possibilities Bit images: 2 1 brightness values 2 brightness values images: 2 6 brightness values 64 brightness values images: 2 7 brightness values 128 brightness values Byte images: 2 8 brightness values 256 brightness values images: 2 10 brightness values 1,024 brightness values images: 2 12 brightness values 4,096 brightness values Integer images: 2 16 brightness values 65,536 brightness values Real values: A decimal point Mathematical sign Unsigned: All values 0.0 Signed: Negative & positive values Common possibilities Real images: 2 32 brightness values

3 The Structure of Digital Images Visual: Rectangular array of pixels Computer: Linear string of values The Display of Digital Images Single-band image display Black & white Gray scale False color Each gray tone a unique color Multi-band image display True color Blue is displayed as blue Green is displayed as green Red is displayed as red False color infrared Green is displayed as blue Red is displayed as green RIR is displayed as red Single-Band Gray & False Color Single-Band False Color Radar False Color Digital Elevation Model Yuma, AZ: False Color Infrared CIR Display IR1 Spectral Band Red Spectral Band Grn Spectral Band

4 Three Image Characteristics Scale (Imaged size / Actual size) Small, medium, large > 1:500,000 to < 1:50,000 Variations Representative fraction (RF): 1:1,000,000 or 1:24,000 Graphical scale: Kilometers Brightness Variations from black to white Digital number values from 0 to 255 (8 bits / pixel) Contrast ratio (CR) Brightest image area / Darkest image area High = 4.5 Medium = 2.5 Low = 1.5 Three More Image Characteristics Resolving power An instrument s ability to form separate images High vs. low quality optics Rigidity of supporting framework Spatial resolution A medium s ability to record separate images High vs. low resolution film Grain size Large vs. small CCD s Manufacturing techniques Angular resolving power Radians used: α = L / r rad 1 radian = 57.3 Still More Image Characteristics Detectability Something is there Recognizability I know what is there Signatures All characteristics that determine recognizability Baseball diamond, golf courses, rivers Light industry, commercial, apartments Texture Closely spaced variations in brightness & color Human Vision Resolving power Determined by largest receptor cells in the fovea 1 / 5,000 = radians = Three-color + intensity vision The human eye has two types of sensors Rods Sensitive to brightness = Intensity Cones Sensitive to colors = Red, green & blue California condors also have two types of sensors All those that human eyes have + Reflected infrared that results in 4-color vision!!! Human (Trichromatic) Vision Bird (Tetrachromatic) Vision

5 Remote Sensing Systems Framing systems Instantaneous acquisition over entire covered area Cameras, CCD arrays Image geometry closely corresponds to Earth geometry Remote Sensing Systems Scanning systems Cross-track scanners Passive Oscillating mirror Horizontal axis Rotating mirror Horizontal axis Along-track scanners Passive No moving mirrors Pushbroom scanners Side-looking scanners Active Radar Sonar Sound rather than EMR Remote Sensing Systems Crucial energy characteristics Energy flux Energy reflected or emitted per unit area of terrain Altitude Higher altitude means less energy received Energy received at sensor is proportional to H -2 Spectral bandwidth of detectors Narrow spectral sensitivity means less energy received Instantaneous field of view (IFOV) Higher spatial resolution means less energy received Dwell time Longer dwell time means more energy received Multispectral Imaging Systems Multispectral still cameras Multiple still cameras, each w/an appropriate filter Panchromatic (B / W) film CCD arrays Multispectral video cameras Multiple video cameras, each w/an appropriate filter Multiple split & X prisms Daedalus AMS airborne scanner system 10 spectral bands from UV to RIR Natural color images can be produced Multispectral Video System Daedalus AMS Multispectral Scanner

6 Sample Multispectral Curves Hyperspectral Imaging Systems As many as 255 spectral bands Received energy issues Substantial correlation between similar bands Looking for unusual/unique spectral features Reflectance maxima Absorption minima Operational Hyperspectral Systems GER Hyperspectral Scanner 63 spectral bands 24 spectral bands 0.50 to 1.00 µm Grn to NIR 7 spectral bands 1.00 to 2.00 µm Water vapor 32 spectral bands 2.00 to 2.50 µm MIR Daedalus 102 spectral bands Visible NIR, MIR & TIR Geoscan 24 recorded bands from 46 available bands Experimental Imaging Spectrometers AVIRIS (Airborne Vis/IR Imaging Spectrometer) 224 spectral bands 10 nm bandwidth Visible NIR & MIR Needs frequent recalibration (~ every 30 minutes) Hyperspectral Reflectance Curves Laboratory spectra Obtained under optimal conditions Pure mineral specimens Calibrated spectrometer & illumination source No atmospheric effects Hyperspectral scanner spectra Obtained under sub-optimal conditions Impure mineral specimens & mixed pixels No two scanners yield identical reflectance curves All scanners detect major absorption features Sensor bandwidth is a major factor Surface Optics Corporation 170VP Portable scanning hyperspectral camera SOC-170VP weighs only 6.6 pounds Based on a diffraction grating rather than a prism Spectral range of 400 to 1,000 nanometers Radiometric resolution of 12 bits per pixel

7 Drones: Fixed-Wing Drones: Rotary-Wing Drones: Lighter-Than-Air Miniturized Imagers: Visible Miniturized Imagers: Reflected IR Miniturized Imagers: Radiated IR iphone 5S & FLIROne