Detailed ExplanationGELE · Photogrammetry & CartographyReal content

GELE Photogrammetry & Cartography — Stereoscopy, DEM and OrthophotoDetailed Explanation

Summary Concepts Notes Revision Deep Dive Slides Flashcards Mistakes Cheat Mnemonics Templates

Want to really understand Stereoscopy, DEM and Orthophoto before tackling GELE Photogrammetry & Cartography questions? This detailed explanation breaks down every key concept, shows you why it matters for the GELE 2026, and walks through the reasoning Professional Regulation Commission (PRC) — Board of Geodetic Engineering expects on high-difficulty questions.

Exam context

For the Geodetic Engineer Licensure Examination, Professional Regulation Commission (PRC) — Board of Geodetic Engineering tests Photogrammetry & Cartography under a "Core" label, with Stereoscopy, DEM and Orthophoto in the 3rd slot across 6 chapters. GELE candidates must clear the 70% weighted average, no sub-test below 50% cut on the 2026 paper, which draws about a meaningful share of Photogrammetry & Cartography questions. Date to watch: September 2026.

Stereoscopy, DEM and Orthophoto - Detailed Explanation

Photogrammetry is the science of obtaining reliable measurements from photographs. Among its most powerful capabilities is the extraction of three-dimensional terrain information from overlapping aerial or satellite images — a process rooted in stereoscopy. Once a three-dimensional model of the terrain is established in the form of a Digital Elevation Model (DEM), it becomes possible to remove the geometric distortions inherent in raw photographs and produce orthophotos: images that have the metric accuracy of a map combined with the visual richness of a photograph. For PRC Geodetic Engineer board examinees, this trio of concepts — stereoscopy, DEM, and orthophoto — appears consistently in the Photogrammetry and Cartography subject area. Mastery requires understanding not only the definitions but also the quantitative relationships (base-height ratio, overlap geometry, parallax) and the practical workflow from raw imagery to a usable map product. This chapter builds on the parallax and relief-displacement concepts covered in earlier photogrammetry chapters and extends them into the production of 3-D terrain data and map-accurate imagery.

Concepts

Stereoscopy and the Stereo Model

Stereoscopy is the technique of perceiving depth (three dimensions) from two slightly different two-dimensional images of the same scene taken from different camera positions. This mimics human binocular vision: each eye sees a slightly different view, and the brain fuses both into a single 3-D perception. In aerial photogrammetry, successive photographs along a flight strip are taken with approximately 60% forward overlap (also called end lap). This means each ground object appears in at least two consecutive photos. When a photogrammetrist views these two photos — called a stereopair — through a stereoscope (optical instrument) or in digital photogrammetric software, the brain or the algorithm fuses the two images into a floating 3-D stereo model. The fundamental measurement in the stereo model is parallax (x-parallax), which is the difference in image position of the same ground point between the left and right photos (see Chapter 2). Differences in parallax across the stereo model directly encode terrain elevation: points at higher elevation have greater parallax than lower points. This is why parallax measurements in the stereo model are the basis for extracting terrain heights. Key geometry: - Air-base (B): the horizontal distance between the two exposure stations (camera positions) in the air. - Flying height (H): the height of the aircraft above the datum (usually Mean Sea Level). - Principal distance / focal length (f): the camera's focal length. - Forward overlap (p%): the percentage of a photo that is duplicated in the next photo along the flight line. The air-base is directly related to overlap: B = (1 − p/100) × L where L is the ground coverage of one photo along the flight direction. For example, with 60% overlap: B = (1 − 0.60) × L = 0.40 × L. The stereo model has a definite geometric strength for height determination, which is controlled by the base-height ratio B/H. A larger ratio means the two camera positions are farther apart relative to the flying height, giving a stronger (more sensitive) stereo geometry for measuring heights.

Examples

The air-base is 40% of the ground coverage because the remaining 60% is overlapped with the previous photo. This value is used to compute the base-height ratio.

Scenario

A flight over Metro Manila uses a camera with a 153 mm focal length at a scale of 1:10,000 and a 230 mm × 230 mm film format with 60% forward overlap. Find the air-base.

Solution

Step 1 — Ground coverage along flight direction: L = format size × scale denominator L = 0.230 m × 10,000 = 2,300 m Step 2 — Air-base: B = (1 − 0.60) × 2,300 B = 0.40 × 2,300 B = 920 m

A B/H of 0.50 is acceptable for photogrammetric height measurement. The theoretical ideal for strong geometry is B/H ≈ 0.6 (corresponding to 60% overlap on a standard format). Values below 0.3 give weak height determination.

Scenario

Two successive exposures over Quezon City are 850 m apart (air-base). The flying height above datum is 1,700 m. What is the base-height ratio?

Solution

B/H = 850 / 1700 = 0.50

Applications

Topographic mapping of Philippine mountain ranges (Cordillera, Sierra Madre) where ground access is difficult.
Production of the Philippine Topographic Map Series at 1:50,000 and 1:10,000 scales using aerial stereopairs.
Disaster risk assessment — stereo DEMs reveal landslide-prone slopes in Bicol and Eastern Visayas.
Urban 3-D city modelling for Metro Manila using drone-based stereo imagery.
Cadastral surveys under PD 1529 may use stereophotogrammetry to supplement ground surveys in remote areas.

Misconceptions

MISCONCEPTION: '60% overlap means the air-base is 60% of the ground coverage.' CORRECTION: 60% overlap means 60% is SHARED with the next photo, so the air-base is the remaining 40% of ground coverage.
MISCONCEPTION: 'Any two photos from the same flight can form a stereopair.' CORRECTION: Only consecutive photos with sufficient overlap (~60%) form a usable stereopair.
MISCONCEPTION: 'Higher overlap always gives better results.' CORRECTION: Very high overlap (>80%) wastes flight resources; very low overlap (<50%) gives weak or no stereoscopic coverage.
MISCONCEPTION: 'The stereo model gives true ground elevations directly.' CORRECTION: The stereo model gives relative heights; absolute elevations require ground control points (GCPs).

Related Concepts

Parallax and x-parallax equations (Chapter 2)
Relief displacement
Ground control points (GCPs)
Aerial triangulation / bundle adjustment
Base-height ratio and vertical exaggeration

Common Exam Questions

Example

A 230 mm format camera flies at 1:8,000 scale with 65% overlap. Air-base = (1 − 0.65) × (0.230 × 8,000) = 0.35 × 1,840 = 644 m.

Approach

Identify the photo format, map scale (to get ground coverage), and overlap percentage. Apply B = (1 − p/100) × L.

Question Type

Numerical — Air-base computation

Example

Question: 'Why must aerial photos overlap for photogrammetric height determination?' Answer: Overlap creates a stereopair; the parallax difference between corresponding image points encodes the height difference between ground points.

Approach

Explain that two views from different positions create parallax differences that encode elevation, enabling 3-D measurement.

Question Type

Conceptual — Purpose of stereoscopy

Example

B = 1,200 m, H = 2,400 m → B/H = 0.50.

Approach

Divide air-base by flying height. Interpret the result relative to the standard ~0.6.

Question Type

Numerical — Base-height ratio

Key Points To Remember

~60% forward overlap is standard for stereo photogrammetry; ~30% sidelap between adjacent flight strips.
Air-base B = (1 − p/100) × ground coverage along flight direction.
Stereoscopic parallax differences encode terrain elevation differences.
A stereopair consists of exactly two overlapping photos from different camera positions.
The stereo model is the 3-D geometric reconstruction from a stereopair, used to extract elevations.
Digital photogrammetric workstations automate stereo matching to produce dense point clouds.

Base-Height Ratio and Vertical Exaggeration

The base-height ratio (B/H) is one of the most important geometric parameters in stereoscopic photogrammetry. It is defined as: B/H = (air-base) / (flying height above datum) This dimensionless ratio controls two critical aspects of the stereo model: 1. HEIGHT MEASUREMENT STRENGTH (Geometric Strength) A larger B/H means the two camera positions subtend a larger angle to any ground point, making height differences more pronounced as parallax differences. This leads to more precise height determination. Conversely, a small B/H (e.g., 0.1) means the two photos are nearly identical, giving very small parallax differences and poor height accuracy. For standard 60% overlap on a 230 mm format, B/H ≈ 0.40–0.60 is typical and acceptable. 2. VERTICAL EXAGGERATION In the stereo model viewed through a stereoscope, the perceived vertical scale is amplified relative to horizontal scale. This exaggeration factor is approximately: VE ≈ (1/B/H) × (1/magnification) For practical purposes, a larger B/H produces less vertical exaggeration (closer to reality), while a smaller B/H produces more vertical exaggeration (terrain looks steeper than it is). Practical guidance: - B/H < 0.3: Weak stereo, poor height accuracy, avoid if possible. - B/H ≈ 0.4–0.6: Standard range, good balance of stereo coverage and height accuracy. - B/H > 0.8: Strong height accuracy but may cause stereo dead areas (gaps in stereo coverage for large relief). Relationship to overlap: B/H = [(1 − p/100) × f] / H × (H/f) = (1 − p/100) × (ground format / H) Simplified: for a given format and flying height, overlap directly determines B/H. For board exam problems, always express B and H in the same units (metres) before computing the ratio.

Examples

Note that B/H = (1 − p/100) × (L/H) = (1 − p/100) × (format/f) = 0.40 × (0.230/0.152) = 0.40 × 1.513 = 0.605. This shortcut is very useful: B/H depends only on overlap and the format-to-focal-length ratio, not on flying height directly!

Scenario

A photogrammetric survey uses a 152 mm focal length camera flying at H = 3,000 m above datum with 60% forward overlap on a 230 mm format. Compute B/H.

Solution

Step 1 — Scale denominator: Scale = H/f = 3,000 / 0.152 = 19,737 ≈ 1:19,737 → use as scale for ground coverage Step 2 — Ground coverage along flight: L = 0.230 × 19,737 ≈ 4,540 m Step 3 — Air-base: B = (1 − 0.60) × 4,540 = 0.40 × 4,540 = 1,816 m Step 4 — B/H: B/H = 1,816 / 3,000 = 0.605

In practice, if B/H < 0.3, the project specifications should be revised — either reduce overlap or increase the format size — to improve geometric strength for height measurements.

Scenario

A surveyor reports B/H = 0.25 for a photogrammetric project. Should this be a concern?

Solution

Yes. B/H = 0.25 is below the recommended minimum of 0.3. This indicates either very high overlap (>75%) or a narrow format, resulting in weak height determination. Elevation errors will be large.

Applications

Designing photogrammetric flight plans for Philippine topographic mapping projects (NAMRIA).
Specifying camera and flying height parameters to achieve required contour interval accuracy.
Evaluating the geometric quality of existing stereopairs before committing to DEM extraction.
Vertical exaggeration awareness when interpreting stereo models for geological mapping of Philippine fault zones (e.g., Philippine Fault Zone).

Misconceptions

MISCONCEPTION: 'B/H has units of metres per metre.' CORRECTION: B/H is dimensionless — it is a pure ratio.
MISCONCEPTION: 'A larger B/H always exaggerates terrain more.' CORRECTION: It is the SMALLER B/H that causes MORE vertical exaggeration in the stereo model.
MISCONCEPTION: 'B/H depends primarily on flying height.' CORRECTION: B/H = (1-p/100) × (format/f), so it depends on overlap and camera geometry, not directly on flying height.

Related Concepts

Forward overlap and air-base
Parallax measurement and height determination
Flight planning parameters
Vertical accuracy of DEM
Stereo model orientation

Common Exam Questions

Example

60% overlap, 230 mm format, 152 mm focal length: B/H = 0.40 × (230/152) = 0.40 × 1.513 = 0.605.

Approach

Use B/H = (1 − p/100) × (format size / focal length) when flying height is not given but format and focal length are known.

Question Type

Shortcut formula for B/H

Example

Increasing overlap from 60% to 80% reduces B/H from 0.40×(L/H) to 0.20×(L/H) — halving the geometric strength for heights.

Approach

Higher overlap → smaller B (less advance per photo) → smaller B/H → weaker height determination.

Question Type

Effect of overlap on B/H

Key Points To Remember

B/H = air-base ÷ flying height (both in metres; result is dimensionless).
Larger B/H → stronger height measurement + less vertical exaggeration.
Smaller B/H → weaker height measurement + more vertical exaggeration.
Standard 60% overlap gives B/H ≈ 0.4–0.6 for typical formats.
Vertical exaggeration causes terrain to appear steeper than actual — important when interpreting stereo models visually.
B/H is also used in the formula for maximum height that can be mapped without stereo dead areas.

Digital Elevation Model (DEM)

A Digital Elevation Model (DEM) is a digital representation of the terrain surface, typically expressed as a regular grid (raster) or Triangulated Irregular Network (TIN) of elevation values referenced to a vertical datum. TERMINOLOGY (Critical for Board Exams): - DEM (Digital Elevation Model): Generic term; often used interchangeably with DTM in Philippine practice. - DTM (Digital Terrain Model): Bare-earth model representing the actual ground surface, excluding above-ground features (trees, buildings). - DSM (Digital Surface Model): Represents the top of ALL surfaces, including buildings, vegetation, and infrastructure. DSM ≥ DTM always. PRODUCTION METHODS: 1. Photogrammetric Image Matching: Automated algorithms (semi-global matching, SGM) find corresponding pixels in stereopairs and compute parallax → elevation. Used with aerial and satellite imagery. Output: dense point cloud → gridded DEM. 2. LiDAR (Light Detection and Ranging): Active sensor emits laser pulses; measures return time → precise distance → elevation. Provides both first return (DSM) and last/bare-earth return (DTM). LiDAR gives highest accuracy (±5–15 cm vertical) and is the standard for hazard mapping in the Philippines (Project NOAH, PhilLiDAR). 3. InSAR (Interferometric Synthetic Aperture Radar): Uses phase difference between two SAR images to compute elevation. Works day/night, through clouds — critical for the cloudy Philippine archipelago. SRTM (Shuttle Radar Topography Mission) DEM used as global baseline; 30 m and 90 m resolution available. 4. Ground survey / GPS levelling: Sparse but highly accurate; used to validate and control photogrammetric DEMs. DEM APPLICATIONS: - Contour generation at any desired interval. - Volume and earthwork computation (cut-and-fill for road and dam projects). - Viewshed analysis (telecommunications tower placement, military operations). - Watershed delineation and drainage analysis (flood hazard mapping — critical in typhoon-prone Philippines). - Slope and aspect mapping for land use planning under CA 141 (Public Land Act). - Orthorectification of satellite and aerial imagery (the critical link to orthophotos). - Site suitability analysis for subdivision development under PD 1529. DEM ACCURACY: Vertical accuracy is commonly reported as RMSE (Root Mean Square Error) of spot heights compared to check points measured by GPS or levelling. NAMRIA specifies vertical accuracy standards for topographic maps aligned with international standards (e.g., ASPRS, NSSDA). DEM RESOLUTION: The grid spacing (pixel size) determines how much terrain detail is captured. Typical values: - 1 m: High-resolution drone or LiDAR DEM for engineering projects. - 5–10 m: Airborne LiDAR provincial mapping. - 30 m: SRTM global DEM (freely available, used for reconnaissance). - 90 m: Older SRTM / ASTER GDEM (regional analysis only).

Examples

SRTM (30 m) would be inadequate — its resolution cannot support 5 m contours at 1:10,000. Always match DEM resolution and accuracy to the required map scale and contour interval.

Scenario

A geodetic engineer is preparing a 1:10,000 topographic map of a portion of Nueva Ecija with a contour interval of 5 m. What DEM resolution and production method would be most appropriate?

Solution

For a 1:10,000 map with 5 m contour interval: - DEM resolution: 1–5 m grid spacing (rule of thumb: DEM pixel ≤ contour interval / 2 = 2.5 m; use 1–2 m). - Production method: Airborne LiDAR or dense image matching from large-scale aerial photos (1:8,000 to 1:15,000). - LiDAR is preferred because it penetrates vegetation canopy to give bare-earth elevations needed for accurate contours.

For topographic mapping, cadastral surveys, and flood modelling, the DTM is needed. The DSM is used for telecommunications line-of-sight analysis or forest biomass estimation. LiDAR uniquely provides both from a single flight.

Scenario

Differentiate between the DSM and DTM produced by a LiDAR survey over a forested area of Palawan.

Solution

DSM (Digital Surface Model): Elevation of the top of the forest canopy — the first laser returns from the treetops. Represents the 'envelope' surface seen from above. DTM (Digital Terrain Model): Elevation of the bare ground beneath the forest — derived from last laser returns that penetrated the canopy gaps. Represents the actual terrain. Difference = Canopy Height Model (CHM) = DSM − DTM, which gives tree heights.

Applications

Flood inundation modelling for storm surge hazard mapping in coastal Philippine provinces (Project NOAH).
Road alignment design in the Cordillera — slope and cut-fill volumes from DEM.
Transmission line routing in Luzon — viewshed and terrain clearance from DEM.
Agricultural land classification (A&D vs. forest land) under CA 141 using slope maps derived from DEM.
Seismic hazard assessment along the Philippine Fault Zone — DEM reveals fault scarps and geomorphic indicators.
Watershed management plans required by DENR using DEM-derived drainage networks.

Misconceptions

MISCONCEPTION: 'DEM and DSM are the same thing.' CORRECTION: DEM/DTM = bare ground; DSM = includes buildings and vegetation. They differ in areas with above-ground features.
MISCONCEPTION: 'SRTM is good enough for engineering design.' CORRECTION: SRTM (30 m, ±16 m vertical RMSE) is suitable only for reconnaissance; engineering design needs LiDAR or detailed photogrammetric DEM.
MISCONCEPTION: 'A DEM is just a contour map in digital form.' CORRECTION: A DEM is a gridded elevation dataset; contours are DERIVED from a DEM. The DEM is the primary data; contours are a product.
MISCONCEPTION: 'Orthorectification can be done without a DEM.' CORRECTION: Orthorectification specifically requires a DEM to correct for relief displacement. Without a DEM, only the camera tilt is removed (producing a rectified photo, not an orthophoto).

Related Concepts

Orthorectification and orthophoto production
Contour generation and interpolation
LiDAR point cloud processing
InSAR and radar remote sensing
Ground control points for DEM accuracy
Vertical datum (MSL, ellipsoid) and geoid undulation

Common Exam Questions

Example

Question: 'Which model would a hydraulic engineer use for flood routing?' Answer: DTM (bare-earth), because floodwater flows over the ground, not over building rooftops.

Approach

Memorize: DSM includes all objects (buildings, trees); DTM/DEM = bare earth. DSM ≥ DTM always.

Question Type

DEM vs DSM vs DTM distinction

Example

For mapping a cloudy, forested island in Eastern Visayas: InSAR for regional coverage + LiDAR for high-accuracy local DTM.

Approach

Match sensor to conditions: LiDAR = highest accuracy + canopy penetration; InSAR = cloud cover; Photogrammetry = cost-effective for open terrain.

Question Type

DEM production method selection

Example

Orthorectification requires a DTM (bare-earth) or at minimum a coarse DEM; it does NOT need a DSM.

Approach

For each DEM application (contours, volumes, viewshed, drainage, orthorectification), know the input DEM type required and the output product.

Question Type

DEM application identification

Key Points To Remember

DEM/DTM = bare-earth elevations; DSM = all surface features included.
Three main photogrammetric DEM sources: image matching, LiDAR, InSAR (radar).
LiDAR gives highest vertical accuracy (±5–15 cm); InSAR works through clouds.
DEM is essential input for orthorectification — without it, orthophotos cannot be made.
DEM drives contour generation, volume computation, drainage analysis, and viewshed.
SRTM 30 m DEM is freely available and commonly referenced in Philippine projects.
PhilLiDAR / Project NOAH produced high-resolution LiDAR DEMs for hazard-prone Philippine areas.

Orthophoto and Orthorectification

UNDERSTANDING THE PROBLEM WITH RAW AERIAL PHOTOGRAPHS A raw (uncorrected) aerial photograph is NOT a map. It has two types of geometric distortions: 1. Tilt displacement: Camera tilt from the vertical shifts image positions; corrected by rectification. 2. Relief displacement: Objects at different elevations are displaced radially from the photo nadir; corrected only with a DEM (elevation data). Because scale varies with terrain elevation (objects on hills are closer to the camera and thus appear at a larger scale), distances measured on a raw photo will be wrong — especially in hilly or mountainous terrain. THE ORTHOPHOTO DEFINED An orthophoto (also called an orthophotograph or orthorectified image) is a photograph that has been geometrically corrected (orthorectified) using: - The camera's interior orientation (focal length, principal point, lens distortion). - The camera's exterior orientation (position and attitude at the moment of exposure). - A Digital Elevation Model (DEM) — to account for each pixel's elevation. The result is an image at a uniform, constant scale where every point is in its correct planimetric position — identical to what a vertical map would show, but with photographic detail. An orthophoto CAN be measured like a map. ORTHORECTIFICATION PROCESS (STEP BY STEP): 1. Acquire raw aerial/satellite images with ground control points (GCPs). 2. Perform aerial triangulation / bundle adjustment to determine camera exterior orientation for all photos. 3. Generate or acquire a DEM of the project area. 4. For each output pixel in the orthophoto: a. Project the pixel's ground coordinates and DEM elevation back through the camera model. b. Determine the corresponding location in the raw image. c. Resample (interpolate) the raw image grey value to the output pixel. 5. Repeat for all pixels → one orthophoto per input image. 6. Mosaic individual orthophotos into a seamless orthomosaic. ORTHOMOSAIC When many overlapping orthophotos are merged (mosaicked) into a single seamless image, the result is an orthomosaic (also called a digital orthophoto map, DOM). This is the standard deliverable for cadastral, topographic, and land administration mapping projects in the Philippines. METRIC PROPERTIES OF AN ORTHOPHOTO: - Uniform scale throughout. - Distances, angles, and areas can be measured directly. - Coordinates can be read from the image using the map projection (e.g., PPCS/TM or UTM WGS84). - Can be used as a base layer in GIS for land registration under PD 1529. ORTHOPHOTO vs MAP vs RAW PHOTO: | Property | Raw Photo | Orthophoto | Topographic Map | |-------------------|---------------|----------------|-----------------| | Scale | Variable | Uniform | Uniform | | Measurable | No (general) | Yes | Yes | | Visual detail | High | High | Low (symbolic) | | Relief correction | No | Yes (via DEM) | N/A | | Tilt correction | No | Yes | N/A | RESOLUTION / GROUND SAMPLING DISTANCE (GSD): The spatial resolution of an orthophoto is described by its Ground Sampling Distance (GSD) — the physical size of one pixel on the ground: GSD = pixel size on sensor × (H / f) For example: pixel = 5 µm, H = 1,500 m, f = 150 mm: GSD = 0.005 mm × (1,500,000 mm / 150 mm) = 0.005 × 10,000 = 50 mm = 5 cm NAMRIA specifies GSD requirements for standard orthophoto products (e.g., GSD ≤ 50 cm for 1:10,000 orthophotos).

Examples

For any measurement or legal mapping purpose, always use an orthophoto, never a raw photo. This distinction is fundamental and frequently tested in the PRC board exam.

Scenario

A real estate developer needs to measure lot areas on aerial imagery of a subdivision in Cavite. The geodetic engineer presents two options: (A) raw aerial photo, (B) orthophoto. Which should be used and why?

Solution

Use Option B: Orthophoto. Reason: The raw aerial photo has variable scale due to terrain relief and camera tilt. Distances and areas measured on it will be geometrically incorrect — especially near elevation changes. The orthophoto has been orthorectified using a DEM, giving uniform scale throughout. All planimetric measurements (distances, bearings, areas) from the orthophoto are geometrically correct and legally defensible for cadastral purposes under PD 1529.

A 10 cm GSD orthophoto can show features as small as 0.3–0.5 m (rule of thumb: minimum detectable object ≈ 3–5 pixels). This is suitable for 1:2,000 to 1:5,000 scale cadastral mapping.

Scenario

An orthophoto project over a hilly area of Benguet uses a camera with a 5 µm pixel size and 150 mm focal length flying at 3,000 m above the average terrain. Compute the GSD.

Solution

GSD = pixel size × (H / f) Convert to consistent units: pixel = 0.000005 m, H = 3,000 m, f = 0.150 m GSD = 0.000005 × (3,000 / 0.150) GSD = 0.000005 × 20,000 GSD = 0.10 m = 10 cm

This is a nuanced exam question. The key principle: orthorectification always uses a DEM. The accuracy IMPACT of omitting the DEM decreases as terrain becomes flatter. But the process of orthorectification by definition includes DEM correction.

Scenario

An orthophoto is produced over a flat coastal plain in Pampanga. A colleague argues that since terrain relief is minimal, a simple rectification (without DEM) would produce an equally accurate result. Is this correct?

Solution

Partially correct but technically incomplete. In very flat terrain, relief displacement is negligible, so the difference between rectified photo and orthophoto is minimal. However, even in flat terrain, small elevation differences (e.g., dike embankments 1–2 m high, road fillings) can cause measurable displacement for large-scale products (1:1,000). For production of a proper orthophoto by definition, a DEM is still used — even a simple flat-plane DEM. The argument is more valid for reconnaissance-scale products but not for high-accuracy cadastral work.

Applications

NAMRIA's production of digital orthophoto maps (DOM) as the primary base map for Philippine land administration.
Cadastral survey base maps for PD 1529 (Property Registration Decree) lot delineation and titling.
Land use mapping and verification of agricultural land under CA 141 (Commonwealth Act No. 141 — Public Land Act).
Post-disaster rapid mapping after typhoons or earthquakes using drone-based orthomosaics.
Infrastructure asset inventory (road network, bridges) using orthophotos as a base layer.
Verification of foreshore areas and reclamation extents under DENR and PRA jurisdiction.
Subdivision plan preparation and checking for compliance with BP 220 and PD 957.

Misconceptions

MISCONCEPTION: 'An aerial photo is the same as an orthophoto.' CORRECTION: A raw aerial photo has variable scale due to tilt and relief; an orthophoto has been geometrically corrected to uniform scale.
MISCONCEPTION: 'Orthorectification is just rotating and scaling the photo.' CORRECTION: Orthorectification is a pixel-by-pixel correction using the camera model and DEM elevation at each point.
MISCONCEPTION: 'Rectification = orthorectification.' CORRECTION: Rectification removes only camera tilt (using a flat plane); orthorectification removes tilt AND relief displacement (using a DEM).
MISCONCEPTION: 'A high-resolution orthophoto automatically has high positional accuracy.' CORRECTION: High resolution (small GSD) means fine detail, but positional accuracy depends on GCP quality, DEM accuracy, and aerial triangulation — not GSD alone.
MISCONCEPTION: 'Orthomosaics have no geometric errors.' CORRECTION: Orthomosaics still contain errors from DEM inaccuracies, GCP errors, and resampling — they meet accuracy standards within specified tolerances.

Related Concepts

Relief displacement
Image rectification
Ground Control Points (GCPs)
Aerial triangulation and bundle adjustment
GSD and image scale
Orthomosaic production workflow
Map accuracy standards (NSSDA, NAMRIA)

Common Exam Questions

Example

Question: 'What distinguishes an orthophoto from a rectified photograph?' Answer: An orthophoto uses a DEM to correct for relief displacement in addition to tilt correction; a rectified photo corrects only for camera tilt.

Approach

Rectification removes tilt only; orthorectification removes tilt AND relief displacement. Only the latter requires a DEM.

Question Type

Orthophoto vs rectified photo distinction

Example

Pixel = 6 µm = 0.006 mm; H = 2,000 m = 2,000,000 mm; f = 200 mm. GSD = 0.006 × (2,000,000/200) = 0.006 × 10,000 = 60 mm = 6 cm.

Approach

GSD = pixel size (m) × (flying height / focal length). Keep units consistent (all in metres or mm).

Question Type

GSD computation

Example

Without a DEM, the orthorectification algorithm assumes all ground points lie on a flat reference plane — leaving residual displacement errors proportional to terrain height variations.

Approach

Explain relief displacement: objects at different heights are displaced by d = r × ΔH / H from nadir. The DEM provides ΔH for every pixel, allowing pixel-by-pixel correction.

Question Type

Why is a DEM needed for orthorectification?

Key Points To Remember

Raw photo ≠ map: scale varies with relief and tilt — do NOT measure distances on a raw photo.
Orthophoto corrects BOTH tilt AND relief displacement using a DEM + camera model.
Rectification alone (without DEM) corrects only tilt, not relief displacement — NOT an orthophoto.
Orthophoto has uniform scale and can be measured like a map while retaining photographic detail.
Orthorectification requires: raw image + camera interior/exterior orientation + DEM.
Orthomosaic = multiple orthophotos merged into a seamless image — standard deliverable.
GSD = ground pixel size; smaller GSD = higher resolution orthophoto.
Orthophotos are used as base maps for PD 1529 land registration and cadastral surveys.

Practice Problems

The shortcut B/H = (1 − p/100) × (format/f) is extremely useful for board exam problems where H may not be given. Always verify that B/H is in the range 0.3–0.8 for acceptable stereo geometry. Here, 0.605 is ideal for photogrammetric height determination.

Problem

PROBLEM 1 — Air-base and B/H ratio Aerial photos are taken over a portion of the Cagayan Valley with the following specifications: - Camera focal length: f = 152 mm - Film/sensor format: 230 mm × 230 mm - Flying height above datum: H = 2,280 m - Forward overlap: 60% (a) Compute the photo scale. (b) Compute the ground coverage per photo (along the flight direction). (c) Compute the air-base. (d) Compute the B/H ratio.

Solution

(a) Photo scale: Scale = f / H = 0.152 / 2,280 = 1/15,000 → Scale is 1:15,000 (b) Ground coverage: L = format size × scale denominator = 0.230 m × 15,000 = 3,450 m (c) Air-base: B = (1 − 0.60) × L = 0.40 × 3,450 = 1,380 m (d) B/H ratio: B/H = 1,380 / 2,280 = 0.605 Alternate check using shortcut: B/H = (1 − p/100) × (format/f) = 0.40 × (230/152) = 0.40 × 1.513 = 0.605 ✓

A B/H of 0.537 is within the acceptable range. The 65% overlap reduces the air-base compared to 60% overlap (which would give B/H = 0.605), slightly reducing height-measurement strength but increasing redundancy in stereo coverage — a common trade-off in UAV surveys.

Problem

PROBLEM 2 — Air-base from overlap and scale A UAV survey over a portion of Leyte produces aerial photos at 1:3,000 scale using a 230 mm equivalent format with 65% forward overlap. Find: (a) Ground coverage along flight direction (b) Air-base (c) B/H if flying height is 450 m

Solution

(a) Ground coverage: L = 0.230 m × 3,000 = 690 m (b) Air-base: B = (1 − 0.65) × 690 = 0.35 × 690 = 241.5 m (c) B/H ratio: B/H = 241.5 / 450 = 0.537 Note: Focal length check → f = H / scale_denominator = 450 / 3,000 = 0.15 m = 150 mm Shortcut check: B/H = 0.35 × (230/150) = 0.35 × 1.533 = 0.537 ✓

A 13.6 cm GSD means each pixel covers a 13.6 cm × 13.6 cm area on the ground. This is suitable for medium-scale topographic mapping (1:10,000 to 1:25,000). For cadastral work at 1:1,000, a GSD of ≤5 cm would be required, necessitating a lower flying height or a larger-format sensor.

Problem

PROBLEM 3 — GSD computation A digital aerial camera with a sensor pixel size of 6.8 µm and focal length of 100 mm is flown at 2,000 m above mean terrain. Compute: (a) The Ground Sampling Distance (GSD) (b) The approximate photo scale equivalent

Solution

(a) GSD: GSD = pixel size × (H / f) Convert: pixel = 6.8 µm = 0.0000068 m; H = 2,000 m; f = 0.100 m GSD = 0.0000068 × (2,000 / 0.100) GSD = 0.0000068 × 20,000 GSD = 0.136 m ≈ 13.6 cm (b) Photo scale equivalent: Scale = f / H = 0.100 / 2,000 = 1/20,000 → Scale is 1:20,000

This type of scenario-based problem tests integrated knowledge. Always link DEM resolution to required map scale/contour interval, production method to terrain conditions (vegetation, cloud cover, accuracy), and model type (DTM vs DSM) to the intended application.

Problem

PROBLEM 4 — DEM application scenario A geodetic engineer is tasked to produce a 1:5,000 topographic map of a proposed dam site in Isabela with a 2 m contour interval. The project is in a forested river valley. (a) What minimum DEM resolution is appropriate? (b) What DEM production method is recommended and why? (c) Should the engineer use a DSM or DTM? Explain.

Solution

(a) DEM resolution: Rule: DEM pixel ≤ CI/2 to CI (where CI = contour interval) DEM pixel ≤ 2/2 = 1 m minimum Recommended: 0.5–1.0 m grid spacing (b) Recommended method: Airborne LiDAR Reasons: - Forested terrain: LiDAR laser pulses penetrate canopy gaps to reach bare ground - Required accuracy: LiDAR provides ±10–15 cm vertical RMSE (sufficient for 2 m CI) - Photogrammetric image matching cannot penetrate forest canopy to give bare-earth elevations - InSAR is too coarse (30 m) for a 1:5,000 product (c) Use DTM (bare-earth model): - The topographic map must show ground terrain, not treetops - DTM = last laser returns from ground; DSM = first returns from canopy - For flood routing in the river valley, water flows on the ground → DTM - DSM would inflate elevations in forested areas by the canopy height (5–30 m), creating completely wrong contours

A 6.9% error in raw photo measurement is significant — this would translate to a 69 m error in a 1 km road, which is unacceptable for engineering design. This problem clearly illustrates why orthophotos are required for any metric use of imagery, as mandated by surveying and mapping standards.

Problem

PROBLEM 5 — Orthophoto concept and computation A planner measures a road segment on a raw aerial photo of hilly terrain in Batangas and obtains a length of 2.3 km. The same road measured on a georeferenced orthophoto gives 2.5 km. The GPS measurement of the road is 2.47 km. (a) Which measurement is most reliable for engineering use? (b) Calculate the percentage error of the raw photo measurement relative to the GPS value. (c) What caused the discrepancy in the raw photo measurement?

Solution

(a) Most reliable for engineering use: Orthophoto measurement (2.5 km) - Orthophoto has uniform scale and corrected geometry (tilt + relief removed) - GPS gives ground truth: 2.47 km - Orthophoto error: |(2.5 − 2.47)/2.47| × 100 = 1.2% — acceptable within typical orthophoto accuracy - Raw photo error: |(2.3 − 2.47)/2.47| × 100 = 6.9% — unacceptable for engineering (b) Percentage error of raw photo vs GPS: Error = |(2.3 − 2.47)| / 2.47 × 100 Error = 0.17 / 2.47 × 100 Error = 6.88% ≈ 6.9% (c) Cause of raw photo discrepancy: Relief displacement: In hilly terrain, the camera is closer to hilltops than to valley floors, so hilltops are imaged at larger scale. This distorts distances measured along slopes. Additionally, any camera tilt shifts image positions. Both effects make raw photo measurements unreliable.

This comprehensive problem integrates all three chapter topics: stereo geometry (B, B/H), flight planning (line spacing), and orthophoto production requirements. Board exam problems often combine multiple concepts in a single scenario — practice integrating knowledge across topics.

Problem

PROBLEM 6 — Integrated stereo-DEM-orthophoto workflow A photogrammetric project over Pangasinan uses the following specifications: - Photo scale: 1:12,000; format: 230 mm × 230 mm; f = 152 mm - Forward overlap: 60%; sidelap: 30% - Flying height: H = 1,824 m above MSL Compute: (a) Air-base (B) (b) B/H ratio (c) Width between flight lines (ground coverage × (1 − sidelap)) (d) What additional data is needed to produce orthophotos from these images?

Solution

(a) Ground coverage (along flight): L = 0.230 × 12,000 = 2,760 m Air-base: B = (1 − 0.60) × 2,760 = 0.40 × 2,760 = 1,104 m (b) B/H ratio: B/H = 1,104 / 1,824 = 0.605 (c) Width between flight lines: W_ground = 0.230 × 12,000 = 2,760 m (same format, same scale) Line spacing = (1 − 0.30) × 2,760 = 0.70 × 2,760 = 1,932 m (d) Additional data needed for orthophoto production: 1. Ground Control Points (GCPs) — surveyed with GPS to establish absolute orientation 2. Camera calibration certificate — interior orientation (f, principal point, lens distortion) 3. Digital Elevation Model (DEM) — to correct for relief displacement pixel by pixel 4. Aerial triangulation results — exterior orientation (X₀, Y₀, Z₀, ω, φ, κ) for each photo

Exam Preparation Tips

MEMORIZE THE B/H SHORTCUT: B/H = (1 − p/100) × (format/f). This eliminates the need to compute flying height separately when format and focal length are given. Verify: 60% overlap, 230 mm format, 152 mm focal length → B/H = 0.40 × 1.513 = 0.605.
DISTINGUISH DEM/DTM/DSM CLEARLY: In board exams, a single wrong term can mean a wrong answer. DEM and DTM = bare earth. DSM = includes buildings and trees. LiDAR gives both; photogrammetric image matching typically gives DSM in vegetated areas.
RAW PHOTO ≠ ORTHOPHOTO: This is the most frequently tested concept. Raw photo has variable scale (relief displacement + tilt). Orthophoto has uniform scale (corrected by DEM + camera model). You CANNOT measure distances reliably on a raw photo.
RECTIFICATION vs ORTHORECTIFICATION: Rectification removes camera TILT only (using a flat ground plane). Orthorectification removes BOTH tilt AND relief displacement (using a DEM). Only orthorectification produces a true orthophoto.
LEARN THE DEM PRODUCTION METHODS: For board exam multiple-choice, know: (1) Image matching from stereopairs — standard photogrammetry; (2) LiDAR — highest accuracy, penetrates canopy; (3) InSAR — works through clouds, covers large areas. The Philippines uses all three: SRTM InSAR for baseline, PhilLiDAR for hazard mapping.
OVERLAP NUMBERS: Forward (end) overlap ≈ 60% for stereo. Sidelap between strips ≈ 30%. These are standard values — expect board exam problems to test what happens when these change (B/H, line spacing, number of stereomodels).
GSD FORMULA: GSD = pixel_size × (H/f). Convert everything to the same unit (metres is safest). A smaller GSD means a higher-resolution, larger-scale orthophoto.
PHILIPPINE CONTEXT: NAMRIA is the official mapping agency. PD 1529 governs land registration. CA 141 governs public lands. Orthophotos and DEMs produced by NAMRIA or accredited geodetic engineers underpin these legal frameworks. Know the connection.
PRACTICE PERCENTAGE ERRORS: Board exams love comparison problems (raw photo vs orthophoto vs GPS). Always compute percentage error as: |measured − true| / true × 100%.
VERTICAL EXAGGERATION DIRECTION: A SMALLER B/H = MORE vertical exaggeration (terrain appears steeper). A LARGER B/H = LESS exaggeration (closer to true proportions). Many students get this backwards — practice with examples.

Loading diagram…

In summary

Stereoscopy, Digital Elevation Models, and Orthophotos form an integrated workflow that is central to modern photogrammetric mapping — and central to the PRC Geodetic Engineer Licensure Examination in Photogrammetry and Cartography. The chain is logical and sequential: overlapping stereopairs create the three-dimensional geometric basis for measuring terrain elevation; those elevations are organised into a DEM; and the DEM enables orthorectification of raw imagery into geometrically correct orthophotos that can be measured and used like maps. For board exam success, internalize three core distinctions: (1) Raw photo ≠ orthophoto — scale varies with relief on a raw photo; (2) Rectification ≠ orthorectification — only orthorectification uses a DEM to correct relief displacement; (3) DEM/DTM ≠ DSM — bare earth versus all-surface elevation models serve different purposes. Master the quantitative formulas: B/H = (1 − p/100) × (format/f) for the base-height ratio, and GSD = pixel_size × (H/f) for image resolution. These appear regularly as direct computation questions. In the Philippine context, these technologies underpin NAMRIA's national mapping program, the PhilLiDAR/Project NOAH hazard mapping initiative, and cadastral and land registration work under PD 1529. A practising geodetic engineer in the Philippines will regularly specify, produce, or quality-check DEMs and orthophotos — so this knowledge is not merely for passing an exam, but for professional competence in service to the Filipino public and the national mapping mission.

Previous chapter

Scale, Relief Displacement and Parallax

Next chapter

Map Projections

Ready to practise for the GELE 2026?

Super Tutor's AI review plan adapts to your weak areas and builds a weekly practice schedule around your target GELE exam date.

Start free See the full GELE guide GELE preparation guide

← Back to Photogrammetry & Cartography hub