Skip to content
CampusPrep
AD-803 (A) · AI for Remote Sensing/Quick Revision Short Notes

AI for Remote Sensing (AD-803 (A)) - Unit 1 Short Notes

How unit 1 is examined

This unit covers what remote sensing is, the physics it rests on, the basics of AI, and how AI is combined with remote-sensing data; no topic was asked in the supplied papers, so each is short but complete.

Definition of Remote sensing

<span style="display:inline-block;padding:.16em .6em;border:1.5px solid currentColor;border-radius:999px;font-size:.68em;font-weight:700;letter-spacing:.06em;text-transform:uppercase;opacity:.75">Not asked since 2022</span>

Definition. <mark>Remote sensing is the science and art of obtaining information about an object, area or phenomenon without physical contact, by recording the electromagnetic radiation it reflects or emits with a sensor.</mark>

Key points.

  1. The sensor is carried on a platform such as a satellite, aircraft or drone, away from the target.
  2. The information is stored as images, where each pixel holds a measured brightness value.
  3. The main parts are the energy source, the atmosphere, the target, the sensor, the ground station and the user.
  4. It gives repeated, large-area, objective coverage, including places that are unsafe or hard to reach.
  5. Passive remote sensing records natural energy, usually reflected sunlight, so it works only in daylight for optical bands; active remote sensing carries its own source, so it works day and night and through cloud.
  6. The information is used for mapping, monitoring and decision making in agriculture, forestry, water, urban planning and disaster management.

Diagram. The remote-sensing process.

<figure class="ds-fig" style="margin:1.4rem 0;overflow-x:auto"><svg xmlns="http://www.w3.org/2000/svg" id="dsfig-u1-01" viewBox="0 0 603 166" width="603" height="166" role="img" aria-label="Sun = energy source, Atm = atmosphere, Tgt = target, Sen = sensor on platform, Gnd = ground station, User = analyst"><style>#dsfig-u1-01 .e{stroke:#454C5A;stroke-width:1.4;fill:none}#dsfig-u1-01 .e.hi{stroke:#2340B8;stroke-width:2.6}#dsfig-u1-01 .n{fill:#FFFFFF;stroke:#16181D;stroke-width:1.4}#dsfig-u1-01 .n.hi{fill:#E3E9FC;stroke:#2340B8;stroke-width:2.2}#dsfig-u1-01 .n.rb-b{fill:#16181D;stroke:#16181D}#dsfig-u1-01 .n.rb-r{fill:#BD3227;stroke:#BD3227}#dsfig-u1-01 text{font-family:"JetBrains Mono",ui-monospace,Menlo,Consolas,monospace;font-size:13px}#dsfig-u1-01 .t{fill:#16181D;font-weight:500}#dsfig-u1-01 .t.inv{fill:#FFFFFF;font-weight:700}#dsfig-u1-01 .kd{stroke:#16181D;stroke-width:1.2}#dsfig-u1-01 .dot{fill:#16181D}#dsfig-u1-01 .ann{fill:#2340B8;font-size:11px;font-weight:700}#dsfig-u1-01 .lbl{fill:#6F7787;font-family:system-ui,-apple-system,sans-serif;font-size:12px;font-weight:700}#dsfig-u1-01 .ptr{fill:#2340B8;font-size:12px;font-weight:700}#dsfig-u1-01 .ah{fill:#454C5A}#dsfig-u1-01 .ah.hi{fill:#2340B8}#dsfig-u1-01 .wl rect{fill:#FFFFFF;stroke:#DCE0E7}#dsfig-u1-01 .wl .t{font-size:12px;font-weight:700}#dsfig-u1-01 .wl.hi rect{fill:#2340B8;stroke:#2340B8}#dsfig-u1-01 .wl.hi .t{fill:#FFFFFF}html.dark #dsfig-u1-01 .e{stroke:#B1B7C3}html.dark #dsfig-u1-01 .e.hi{stroke:#8FA3FF}html.dark #dsfig-u1-01 .n{fill:#161920;stroke:#E6E8ED}html.dark #dsfig-u1-01 .n.hi{fill:#1E2748;stroke:#8FA3FF}html.dark #dsfig-u1-01 .n.rb-b{fill:#E6E8ED;stroke:#E6E8ED}html.dark #dsfig-u1-01 .n.rb-r{fill:#FF7E71;stroke:#FF7E71}html.dark #dsfig-u1-01 .t{fill:#E6E8ED}html.dark #dsfig-u1-01 .t.inv{fill:#0F1115}html.dark #dsfig-u1-01 .kd{stroke:#E6E8ED}html.dark #dsfig-u1-01 .dot{fill:#E6E8ED}html.dark #dsfig-u1-01 .ann{fill:#8FA3FF}html.dark #dsfig-u1-01 .lbl{fill:#858D9C}html.dark #dsfig-u1-01 .ptr{fill:#8FA3FF}html.dark #dsfig-u1-01 .ah{fill:#B1B7C3}html.dark #dsfig-u1-01 .ah.hi{fill:#8FA3FF}html.dark #dsfig-u1-01 .wl rect{fill:#161920;stroke:#2A2E37}html.dark #dsfig-u1-01 .wl.hi rect{fill:#8FA3FF;stroke:#8FA3FF}html.dark #dsfig-u1-01 .wl.hi .t{fill:#0F1115}</style><defs><marker id="ah1" viewBox="0 0 10 10" refX="9" refY="5" markerWidth="7" markerHeight="7" orient="auto-start-reverse"><path class="ah" d="M0,1 L9,5 L0,9 z"/></marker><marker id="ahh1" viewBox="0 0 10 10" refX="9" refY="5" markerWidth="7" markerHeight="7" orient="auto-start-reverse"><path class="ah hi" d="M0,1 L9,5 L0,9 z"/></marker></defs><path class="e" d="M54.6,113.8 L127.1,53.4" marker-end="url(#ah1)"/><path class="e" d="M157.8,52.2 L230.3,112.6" marker-end="url(#ah1)"/><path class="e" d="M261,113.8 L333.5,53.4" marker-end="url(#ah1)"/><path class="e" d="M364.2,52.2 L436.7,112.6" marker-end="url(#ah1)"/><path class="e" d="M467.4,113.8 L534.5,57.9" marker-end="url(#ah1)"/><circle class="n" cx="40" cy="126" r="18"/><text class="t" x="40" y="126" dy=".35em" text-anchor="middle">Sun</text><circle class="n" cx="143.2" cy="40" r="18"/><text class="t" x="143.2" y="40" dy=".35em" text-anchor="middle">Atm</text><circle class="n" cx="246.4" cy="126" r="18"/><text class="t" x="246.4" y="126" dy=".35em" text-anchor="middle">Tgt</text><circle class="n" cx="349.6" cy="40" r="18"/><text class="t" x="349.6" y="40" dy=".35em" text-anchor="middle">Sen</text><circle class="n" cx="452.8" cy="126" r="18"/><text class="t" x="452.8" y="126" dy=".35em" text-anchor="middle">Gnd</text><rect class="n" x="531" y="25" width="50" height="30" rx="15"/><text class="t" x="556" y="40" dy=".35em" text-anchor="middle">User</text></svg><figcaption style="font-size:.82em;opacity:.72;margin-top:.45rem">Sun = energy source, Atm = atmosphere, Tgt = target, Sen = sensor on platform, Gnd = ground station, User = analyst</figcaption></figure>

Basis Passive Active
Energy source Natural (Sun, Earth's heat) Sensor's own transmitter
Time of use Mostly daytime for optical bands Day and night
Cloud penetration Poor in optical bands Good in microwave bands
Example Landsat, Sentinel-2 Radar (SAR), LiDAR

Principles of Remote Sensing

<span style="display:inline-block;padding:.16em .6em;border:1.5px solid currentColor;border-radius:999px;font-size:.68em;font-weight:700;letter-spacing:.06em;text-transform:uppercase;opacity:.75">Not asked since 2022</span>

Definition. The principle of remote sensing is that every object reflects, absorbs, transmits and emits electromagnetic radiation (EMR) in its own way, so the recorded energy identifies the object.

Formula. $c = \lambda \nu$, where $c = 3\times10^{8}$ m/s, $\lambda$ is wavelength and $\nu$ is frequency.

Key points.

  1. EMR travels as waves at the speed of light, and a longer wavelength means a lower frequency.
  2. Energy from the source (usually the Sun) passes through the atmosphere, which scatters and absorbs some of it.
  3. At the target the energy is reflected, absorbed or transmitted, and the reflected share differs by material.
  4. The plot of reflectance against wavelength is the spectral signature, which lets the classifier tell water, soil and vegetation apart.
  5. The sensor records the energy, which is transmitted to a ground station, processed and interpreted.
  6. Energy that reaches the target obeys $E_I = E_R + E_A + E_T$ (incident equals reflected plus absorbed plus transmitted), and the sensor mainly measures $E_R$.
  7. Healthy vegetation reflects strongly in near-infrared and absorbs red light because of chlorophyll, water absorbs almost all near-infrared, and dry soil reflects gradually more with wavelength.
  8. Atmospheric windows are the wavelength ranges where the atmosphere lets energy pass, and sensors are designed to work inside them.
Region Approximate wavelength Use
Visible 0.4-0.7 $\mu$m Colour, water depth, built-up areas
Near infrared 0.7-1.3 $\mu$m Vegetation health
Thermal infrared 8-14 $\mu$m Surface temperature
Microwave 1 mm-1 m All-weather radar imaging

Example. Green light of wavelength $0.55\ \mu$m has frequency $\nu = c/\lambda = 3\times10^{8}/0.55\times10^{-6} \approx 5.45\times10^{14}$ Hz.

Pitfall: Do not say the sensor measures the object itself; it measures the radiation leaving it, which the atmosphere has already altered.

Introduction to Artificial Intelligence

<span style="display:inline-block;padding:.16em .6em;border:1.5px solid currentColor;border-radius:999px;font-size:.68em;font-weight:700;letter-spacing:.06em;text-transform:uppercase;opacity:.75">Not asked since 2022</span>

Definition. <mark>Artificial Intelligence is the branch of computer science that builds machines able to perform tasks needing human intelligence, such as learning, reasoning, perception and decision making.</mark>

Key points.

  1. Machine learning, a subset of AI, learns patterns from data instead of following fixed rules.
  2. Deep learning, a subset of machine learning, uses many-layered neural networks to learn features automatically.
  3. Learning can be supervised (labelled data), unsupervised (no labels) or reinforcement-based (rewards).
  4. AI performance depends on the quality and quantity of training data.
  5. A neural network is made of layers of connected units; each unit takes weighted inputs, adds them and passes the result through an activation function, and training adjusts the weights to reduce error.
  6. Typical AI tasks are classification, regression, clustering, detection and prediction, and each has a matching family of algorithms.
  7. Convolutional neural networks (CNNs) suit images because small filters slide over the picture and learn edges, textures and shapes.

Diagram. AI, machine learning and deep learning are nested.

<figure class="ds-fig" style="margin:1.4rem 0;overflow-x:auto"><svg xmlns="http://www.w3.org/2000/svg" id="dsfig-u1-02" viewBox="0 0 517 198" width="517" height="198" role="img" aria-label="Deep learning is a subset of machine learning, which is a subset of AI"><style>#dsfig-u1-02 .e{stroke:#454C5A;stroke-width:1.4;fill:none}#dsfig-u1-02 .e.hi{stroke:#2340B8;stroke-width:2.6}#dsfig-u1-02 .n{fill:#FFFFFF;stroke:#16181D;stroke-width:1.4}#dsfig-u1-02 .n.hi{fill:#E3E9FC;stroke:#2340B8;stroke-width:2.2}#dsfig-u1-02 .n.rb-b{fill:#16181D;stroke:#16181D}#dsfig-u1-02 .n.rb-r{fill:#BD3227;stroke:#BD3227}#dsfig-u1-02 text{font-family:"JetBrains Mono",ui-monospace,Menlo,Consolas,monospace;font-size:13px}#dsfig-u1-02 .t{fill:#16181D;font-weight:500}#dsfig-u1-02 .t.inv{fill:#FFFFFF;font-weight:700}#dsfig-u1-02 .kd{stroke:#16181D;stroke-width:1.2}#dsfig-u1-02 .dot{fill:#16181D}#dsfig-u1-02 .ann{fill:#2340B8;font-size:11px;font-weight:700}#dsfig-u1-02 .lbl{fill:#6F7787;font-family:system-ui,-apple-system,sans-serif;font-size:12px;font-weight:700}#dsfig-u1-02 .ptr{fill:#2340B8;font-size:12px;font-weight:700}#dsfig-u1-02 .ah{fill:#454C5A}#dsfig-u1-02 .ah.hi{fill:#2340B8}#dsfig-u1-02 .wl rect{fill:#FFFFFF;stroke:#DCE0E7}#dsfig-u1-02 .wl .t{font-size:12px;font-weight:700}#dsfig-u1-02 .wl.hi rect{fill:#2340B8;stroke:#2340B8}#dsfig-u1-02 .wl.hi .t{fill:#FFFFFF}html.dark #dsfig-u1-02 .e{stroke:#B1B7C3}html.dark #dsfig-u1-02 .e.hi{stroke:#8FA3FF}html.dark #dsfig-u1-02 .n{fill:#161920;stroke:#E6E8ED}html.dark #dsfig-u1-02 .n.hi{fill:#1E2748;stroke:#8FA3FF}html.dark #dsfig-u1-02 .n.rb-b{fill:#E6E8ED;stroke:#E6E8ED}html.dark #dsfig-u1-02 .n.rb-r{fill:#FF7E71;stroke:#FF7E71}html.dark #dsfig-u1-02 .t{fill:#E6E8ED}html.dark #dsfig-u1-02 .t.inv{fill:#0F1115}html.dark #dsfig-u1-02 .kd{stroke:#E6E8ED}html.dark #dsfig-u1-02 .dot{fill:#E6E8ED}html.dark #dsfig-u1-02 .ann{fill:#8FA3FF}html.dark #dsfig-u1-02 .lbl{fill:#858D9C}html.dark #dsfig-u1-02 .ptr{fill:#8FA3FF}html.dark #dsfig-u1-02 .ah{fill:#B1B7C3}html.dark #dsfig-u1-02 .ah.hi{fill:#8FA3FF}html.dark #dsfig-u1-02 .wl rect{fill:#161920;stroke:#2A2E37}html.dark #dsfig-u1-02 .wl.hi rect{fill:#8FA3FF;stroke:#8FA3FF}html.dark #dsfig-u1-02 .wl.hi .t{fill:#0F1115}</style><defs><marker id="ah2" viewBox="0 0 10 10" refX="9" refY="5" markerWidth="7" markerHeight="7" orient="auto-start-reverse"><path class="ah" d="M0,1 L9,5 L0,9 z"/></marker><marker id="ahh2" viewBox="0 0 10 10" refX="9" refY="5" markerWidth="7" markerHeight="7" orient="auto-start-reverse"><path class="ah hi" d="M0,1 L9,5 L0,9 z"/></marker></defs><line class="e" x1="401.5" y1="39" x2="246.5" y2="103"/><line class="e" x1="246.5" y1="103" x2="91.5" y2="167"/><circle class="n" cx="401.5" cy="39" r="17"/><text class="t" x="401.5" y="39" dy=".35em" text-anchor="middle">AI</text><rect class="n" x="174" y="88" width="145" height="30" rx="8"/><text class="t" x="246.5" y="103" dy=".35em" text-anchor="middle">Machine learning</text><rect class="n" x="30.5" y="152" width="122" height="30" rx="8"/><text class="t" x="91.5" y="167" dy=".35em" text-anchor="middle">Deep learning</text></svg><figcaption style="font-size:.82em;opacity:.72;margin-top:.45rem">Deep learning is a subset of machine learning, which is a subset of AI</figcaption></figure>

Type Data Example in remote sensing
Supervised Labelled Land-cover classification with training samples
Unsupervised Unlabelled K-means clustering of pixels
Reinforcement Reward signal Autonomous drone navigation

Integration of AI in Remote Sensing

<span style="display:inline-block;padding:.16em .6em;border:1.5px solid currentColor;border-radius:999px;font-size:.68em;font-weight:700;letter-spacing:.06em;text-transform:uppercase;opacity:.75">Not asked since 2022</span>

Definition. Integration of AI in remote sensing means applying machine-learning and deep-learning methods to satellite and aerial imagery so that information is extracted automatically and accurately.

Key points.

  1. Remote sensing produces huge volumes of imagery that manual interpretation cannot handle, so AI automates the analysis.
  2. Supervised classifiers and CNNs map land cover; clustering methods group pixels without labels.
  3. AI supports object detection, change detection, and crop and disaster monitoring.
  4. Challenges are the need for labelled samples, high computing cost and models that do not transfer between sensors or regions.
  5. Traditional methods use hand-made features such as NDVI, texture and shape, whereas deep networks learn features directly from the pixels, which usually raises accuracy.
  6. The normal workflow is to collect imagery, pre-process it (radiometric and geometric correction), extract features, train the model, classify or detect, and assess accuracy with a confusion matrix.
  7. Typical applications are land-cover mapping, crop-yield estimation, flood and fire mapping, urban growth monitoring and ship or vehicle detection.
  8. Accuracy is reported as overall accuracy, which is correctly classified pixels divided by total test pixels, along with the kappa coefficient.

Diagram. AI-based remote-sensing workflow.

<figure class="ds-fig" style="margin:1.4rem 0;overflow-x:auto"><svg xmlns="http://www.w3.org/2000/svg" id="dsfig-u1-03" viewBox="0 0 603 166" width="603" height="166" role="img" aria-label="Img = satellite image, Pre = pre-processing, Feat = feature extraction, Mod = trained AI model, Map = output map with accuracy check"><style>#dsfig-u1-03 .e{stroke:#454C5A;stroke-width:1.4;fill:none}#dsfig-u1-03 .e.hi{stroke:#2340B8;stroke-width:2.6}#dsfig-u1-03 .n{fill:#FFFFFF;stroke:#16181D;stroke-width:1.4}#dsfig-u1-03 .n.hi{fill:#E3E9FC;stroke:#2340B8;stroke-width:2.2}#dsfig-u1-03 .n.rb-b{fill:#16181D;stroke:#16181D}#dsfig-u1-03 .n.rb-r{fill:#BD3227;stroke:#BD3227}#dsfig-u1-03 text{font-family:"JetBrains Mono",ui-monospace,Menlo,Consolas,monospace;font-size:13px}#dsfig-u1-03 .t{fill:#16181D;font-weight:500}#dsfig-u1-03 .t.inv{fill:#FFFFFF;font-weight:700}#dsfig-u1-03 .kd{stroke:#16181D;stroke-width:1.2}#dsfig-u1-03 .dot{fill:#16181D}#dsfig-u1-03 .ann{fill:#2340B8;font-size:11px;font-weight:700}#dsfig-u1-03 .lbl{fill:#6F7787;font-family:system-ui,-apple-system,sans-serif;font-size:12px;font-weight:700}#dsfig-u1-03 .ptr{fill:#2340B8;font-size:12px;font-weight:700}#dsfig-u1-03 .ah{fill:#454C5A}#dsfig-u1-03 .ah.hi{fill:#2340B8}#dsfig-u1-03 .wl rect{fill:#FFFFFF;stroke:#DCE0E7}#dsfig-u1-03 .wl .t{font-size:12px;font-weight:700}#dsfig-u1-03 .wl.hi rect{fill:#2340B8;stroke:#2340B8}#dsfig-u1-03 .wl.hi .t{fill:#FFFFFF}html.dark #dsfig-u1-03 .e{stroke:#B1B7C3}html.dark #dsfig-u1-03 .e.hi{stroke:#8FA3FF}html.dark #dsfig-u1-03 .n{fill:#161920;stroke:#E6E8ED}html.dark #dsfig-u1-03 .n.hi{fill:#1E2748;stroke:#8FA3FF}html.dark #dsfig-u1-03 .n.rb-b{fill:#E6E8ED;stroke:#E6E8ED}html.dark #dsfig-u1-03 .n.rb-r{fill:#FF7E71;stroke:#FF7E71}html.dark #dsfig-u1-03 .t{fill:#E6E8ED}html.dark #dsfig-u1-03 .t.inv{fill:#0F1115}html.dark #dsfig-u1-03 .kd{stroke:#E6E8ED}html.dark #dsfig-u1-03 .dot{fill:#E6E8ED}html.dark #dsfig-u1-03 .ann{fill:#8FA3FF}html.dark #dsfig-u1-03 .lbl{fill:#858D9C}html.dark #dsfig-u1-03 .ptr{fill:#8FA3FF}html.dark #dsfig-u1-03 .ah{fill:#B1B7C3}html.dark #dsfig-u1-03 .ah.hi{fill:#8FA3FF}html.dark #dsfig-u1-03 .wl rect{fill:#161920;stroke:#2A2E37}html.dark #dsfig-u1-03 .wl.hi rect{fill:#8FA3FF;stroke:#8FA3FF}html.dark #dsfig-u1-03 .wl.hi .t{fill:#0F1115}</style><defs><marker id="ah3" viewBox="0 0 10 10" refX="9" refY="5" markerWidth="7" markerHeight="7" orient="auto-start-reverse"><path class="ah" d="M0,1 L9,5 L0,9 z"/></marker><marker id="ahh3" viewBox="0 0 10 10" refX="9" refY="5" markerWidth="7" markerHeight="7" orient="auto-start-reverse"><path class="ah hi" d="M0,1 L9,5 L0,9 z"/></marker></defs><path class="e" d="M55.8,115.5 L151.5,51.6" marker-end="url(#ah3)"/><path class="e" d="M184.8,50.5 L274.7,110.5" marker-end="url(#ah3)"/><path class="e" d="M319.6,111.6 L409.5,51.6" marker-end="url(#ah3)"/><path class="e" d="M442.8,50.5 L538.5,114.4" marker-end="url(#ah3)"/><circle class="n" cx="40" cy="126" r="18"/><text class="t" x="40" y="126" dy=".35em" text-anchor="middle">Img</text><circle class="n" cx="169" cy="40" r="18"/><text class="t" x="169" y="40" dy=".35em" text-anchor="middle">Pre</text><rect class="n" x="273" y="111" width="50" height="30" rx="15"/><text class="t" x="298" y="126" dy=".35em" text-anchor="middle">Feat</text><circle class="n" cx="427" cy="40" r="18"/><text class="t" x="427" y="40" dy=".35em" text-anchor="middle">Mod</text><circle class="n" cx="556" cy="126" r="18"/><text class="t" x="556" y="126" dy=".35em" text-anchor="middle">Map</text></svg><figcaption style="font-size:.82em;opacity:.72;margin-top:.45rem">Img = satellite image, Pre = pre-processing, Feat = feature extraction, Mod = trained AI model, Map = output map with accuracy check</figcaption></figure>

Example. An overall accuracy of $\frac{180}{200} = 90\%$ means 180 of 200 test pixels were labelled correctly.

Last-minute revision

  • Remote sensing means obtaining information without physical contact using reflected or emitted EMR.
  • Basic parts: source, atmosphere, target, sensor, ground station, user.
  • $c = \lambda \nu$ with $c = 3\times10^{8}$ m/s.
  • Different materials have different spectral signatures.
  • Atmosphere scatters and absorbs part of the energy.
  • AI means machines performing tasks needing human intelligence.
  • AI contains machine learning, which contains deep learning.
  • Learning types: supervised, unsupervised, reinforcement.
  • AI in remote sensing automates classification, detection and change analysis.
  • Main hurdles: labelled data, computing cost, generalisation.
  • Passive sensors use natural energy, active sensors use their own; radar sees through cloud.
  • Overall accuracy = correct test pixels / total test pixels.

Memory hooks

  • Remote sensing chain: S-A-T-S-G-U (Source, Atmosphere, Target, Sensor, Ground station, User).
  • AI nesting: AI > ML > DL, like Russian dolls.
  • Spectral signature is the "fingerprint" of a material.
  • Passive = Sun's light (daytime), Active = own signal (day and night, radar and LiDAR).
  • Energy balance: incident = reflected + absorbed + transmitted.
  • Vegetation is bright in near-infrared and dark in red.

Coverage checklist

  • Definition of Remote sensing: definition, components, benefits (no past questions).
  • Principles of Remote Sensing: EMR, interaction, spectral signature, $c = \lambda \nu$ (no past questions).
  • Introduction to Artificial Intelligence: definition, ML, DL, learning types (no past questions).
  • Integration of AI in Remote Sensing: uses and challenges (no past questions).
Go to where you left off?

Quick Add to Notes

Save questions, your own notes and screenshots into notes filed by unit. It takes a free account.

Create free account

Have an account? Log in