🎓 Lesson 8
D5
Optimal Overlap & Altitude Settings for Stockpile Volume Accuracy
Optimal overlap and altitude settings are the right amount of image overlap and drone flying height needed to capture clear, accurate 3D scans of stockpiles so volume measurements are trustworthy.
🎯 Learning Objectives
- ✓ Calculate required front-lap and side-lap percentages for a given stockpile geometry and target GSD
- ✓ Design drone flight parameters (altitude, speed, and camera settings) to achieve ≤2% volumetric error on conical stockpiles ≥10,000 m³
- ✓ Analyze point cloud density and mesh resolution to diagnose underlap/overaltitude-induced volume bias
- ✓ Apply ASTM D6782 validation protocols to quantify and report stockpile volume uncertainty
📖 Why This Matters
In mining operations, stockpile volume errors directly impact inventory reconciliation, royalty payments, and production planning. A 5% volume overestimation on a 50,000-ton coal stockpile can cost >$125,000 in misreported value. Poor overlap or excessive altitude causes gaps in point clouds, surface smoothing artifacts, and erroneous cut/fill calculations—making this not just a 'drone setting' issue, but a financial and regulatory accountability requirement.
📘 Core Principles
Photogrammetric accuracy depends on geometric redundancy: front-lap (typically 70–85%) ensures sufficient tie points along flight lines for elevation consistency; side-lap (60–80%) enables stereo triangulation across rows. Altitude controls Ground Sampling Distance (GSD)—the physical size of each pixel on the ground—and inversely affects both spatial resolution and image count. Lower GSD improves detail but increases processing time and risk of occlusion in steep-sided piles. The interplay between overlap, altitude, lens focal length, and sensor resolution determines the achievable volumetric precision per ASTM D6782’s ‘Level 2’ (±2% relative error) benchmark.
📐 GSD–Altitude–Overlap Relationship
Ground Sampling Distance (GSD) defines the smallest resolvable feature and anchors all accuracy decisions. Front-lap and side-lap are calculated from flight geometry and camera parameters to guarantee minimum 3-view redundancy across the stockpile surface.
💡 Worked Example
Problem: A DJI M300 RTK with a 24 mm equivalent lens (full-frame) and 20 MP sensor (5472 × 3648 px) surveys a limestone stockpile. Target GSD = 2 cm. Drone flies at 60 m AGL. Calculate actual GSD and required front-lap to maintain ≥3 overlapping images per surface point.
1.
Step 1: Compute GSD using GSD = (H × IFOV) / f, where H = 60 m, f = 24 mm, and IFOV ≈ sensor pixel pitch = 5.94 µm (derived from 36 mm / 5472 px). So GSD = (60,000 mm × 0.00594 mm) / 24 mm = 14.85 mm ≈ 1.5 cm.
2.
Step 2: For 3-view redundancy, minimum front-lap = 1 − (GSD × image_width_px) / (ground_width_per_image). Ground width = H × 2 × tan(FOV/2); FOV ≈ 84° → half-FOV = 42° → tan(42°) ≈ 0.9 ⇒ ground width ≈ 60 m × 2 × 0.9 = 108 m. Image width at GSD = 5472 px × 0.015 m/px = 82.1 m. So front-lap = 1 − (82.1 / 108) = 0.24 → 24%. But industry requires ≥70% for stockpiles due to texture variability and shadowing — thus 75% is selected.
3.
Step 3: Verify against typical GSD range: For stockpile inventory, 1–3 cm GSD is standard; 1.5 cm meets ASTM D6782 Level 2 requirements when combined with ≥75% front-lap and ≥65% side-lap.
Answer:
The calculated GSD is 1.5 cm, and a front-lap of 75% ensures robust 3-view coverage. This falls within the safe operational range of 70–85% front-lap for industrial stockpiles.
🏗️ Real-World Application
At Newmont’s Boddington Mine (Western Australia), drone surveys of copper-gold ore stockpiles were initially reporting ±4.8% volume variance vs. total station validation. Analysis revealed 58% front-lap and 42 m AGL flights yielding 3.1 cm GSD. After redesigning missions to 78% front-lap, 65% side-lap, and 52 m AGL (achieving 2.2 cm GSD), volumetric repeatability improved to ±1.3%, meeting ASTM D6782 Level 2. Crucially, the revised plan included 15° oblique imagery to resolve pile toe and crest occlusions—demonstrating that overlap and altitude must be paired with acquisition geometry.
✏️ Field Design Exercise
You’re tasked with surveying a 12,000 m³ iron ore stockpile (conical, ~18 m tall, ~32 m base diameter) using a senseFly eBee X (APS-C sensor, 20 MP, 20 mm lens). Target volumetric accuracy: ≤1.8% (ASTM D6782 Level 2). Given max allowable GSD = 2.5 cm, calculate: (a) maximum safe flight altitude (AGL), (b) required front-lap % assuming 80° horizontal FOV, and (c) minimum number of images per stockpile cross-section to ensure ≥3-view redundancy. Show all assumptions and unit conversions.