Checklist: Assessing Photogrammetry and LiDAR Accuracy

Photogrammetry and LiDAR are two powerful tools for creating 3D site models, but their accuracy depends on how you use them. This guide provides a step-by-step checklist to help you ensure the precision of your data, from survey planning to error correction.

Key Points:

Photogrammetry captures detailed textures but requires good lighting and delivers lower point density (5–20 points/m²).
LiDAR works day or night, penetrates vegetation, and delivers high point density (50–200 points/m²).
Use Ground Control Points (GCPs) for better accuracy, especially in large or uneven areas.
Weather and lighting conditions can impact results - LiDAR is more resilient, but both methods are affected by rain and fog.
Regular calibration, overlap adjustments, and real-time monitoring are critical for reliable data.

Quick Comparison:

Feature	LiDAR	Photogrammetry
Point Density	50–200 points/m²	5–20 points/m²
Light Requirements	Works day or night	Needs ambient light
Vegetation	Penetrates dense areas	Limited penetration
Accuracy	±0.4–1.2 inches vertical	±0.4 inches horizontal

This checklist ensures your 3D models are precise, saving time and reducing errors in projects. Let’s dive into the details.

Photogrammetry VS LiDAR - Which Is BETTER?

Survey Planning Steps

Effective planning is the backbone of accurate 3D data collection.

Site Review and Obstacles

Start by reviewing the site to spot anything that might interfere with LiDAR or photogrammetry data. For LiDAR surveys, vegetation density plays a big role in data quality. In areas with light vegetation, like grass, operators can speed up scanning to save time. But when vegetation is dense, slower scanning is necessary to let the laser pulses reach the ground .

For aerial surveys, pinpointing vertical obstacles is crucial for both safety and accuracy. Operators should carefully check for the tallest obstacles between the takeoff spot and the survey area to set the right flight altitudes . After that, secure clear ground references.

GCP Setup Guide

Ground Control Points (GCPs) are critical for precise spatial data. Dan Wawrzyn from Propeller explains, "Ground control is one of the most important elements of an aerial mapping project because it ensures accuracy" .

Site Size	Minimum GCP Requirements
Any size	5 GCPs
Over 20 acres	Add 1 GCP for every 15 extra acres
Elevation changes	Place GCPs at the highest and lowest points

Place markers in flat, open areas with a clear view of the sky .
Spread the points evenly across the survey area, covering the perimeter and center .
Use durable, high-contrast markers that stay in place .
For better point cloud alignment, consider a low-altitude flight pass (98 feet at 6.6 feet/second) over the GCPs .

Once the ground control is set, it's time to focus on equipment preparation.

Equipment Setup

Ensure all sensors, GPS, and IMU components are calibrated to maintain data quality . Adjust LiDAR scan rates and pulse frequencies based on the project’s needs .

Drone batteries should be fully charged before each flight. For storage between surveys, keep them at 60-70% charge (around 3.8 Volts) .

"Regular calibration is key to getting the best lidar results."

Keep an eye on data quality as you work. Use quality assurance tools to confirm that the initial readings are complete, accurate, and consistent .

Data Collection Standards

After pre-flight checks and equipment setup, following standardized procedures for data collection is crucial.

Drone Flight Parameters

When working in hilly areas, maintain a flight altitude of 150–300 feet (45–90 m) and adjust the overlap based on the slope's steepness. Keep in mind that flying at higher altitudes reduces point density .

Parameter	Photogrammetry	LiDAR
Base Station Distance	Within 6.2 miles (10 km)	Within 6.2 miles (10 km)
Altitude Range	150–300 feet	150–300 feet
Point Cloud Density	Depends on speed and overlap	Depends on scan rate and mode
Real-time Monitoring	Required	Required

Real-time monitoring is essential for identifying anomalies during data collection. Keeping the base station within 6.2 miles ensures reliable communication .

Next, it’s important to fine-tune overlap settings for imaging and scanning.

Image and Scan Overlap

For LiDAR, aim for 20–30% side overlap to achieve accurate results .

For photogrammetry, use the following overlap settings:

Front overlap: 60–80% between consecutive images
Side overlap: 60–80% between adjacent flight lines
Overall coverage: Aim for 80–90%

If you're using hybrid systems that combine LiDAR and photographic data, set the photo overlap to about 75% for both front and side coverage while keeping the standard LiDAR overlap settings .

Now, let’s consider how weather and lighting conditions can impact the data.

Weather and Light Conditions

Weather and lighting play a key role in data quality. LiDAR systems are more flexible under varying lighting conditions, while photogrammetry requires consistent and optimal lighting for sharp images .

Key weather factors to keep in mind:

Fog: Scatters light, increasing noise in LiDAR data .
Rain: Can deflect laser beams and block scanners, leading to errors .
Equipment Protection: Systems like YellowScan's (IP55-rated) are designed to resist harsh weather and protect sensitive components .

Avoid data collection during bad weather, as water droplets on scanners can obstruct the view. While LiDAR is more resilient in tough conditions, heavy rain can still impact both LiDAR and photogrammetry. For photogrammetry, plan surveys during early morning or late afternoon when lighting is consistent, and shadows are minimal, ensuring better image quality and accurate feature matching .

sbb-itb-ac6e058

Data Processing Steps

Raw Data Review

Before diving into processing, it's essential to check the foundational elements of your dataset. Here's what to focus on:

Element	What to Check
Coordinate System	Confirm geotags and projection settings are accurate.
Flight Data	Verify altitude, camera settings, and image overlap.
Point Cloud	Evaluate density and ensure full coverage.
Ground Control	Check the distribution and visibility of ground control points.

Once these basics are confirmed, you can move on to aligning your data with precision.

Data Alignment Methods

Start with a two-step registration process: coarse first, then fine-tune. For coarse alignment, use feature-based approaches like point, line, or surface matching. After that, refine the alignment with techniques such as ICP (Iterative Closest Point), RANSAC, or NDT (Normal Distributions Transform), depending on how well the initial alignment turned out .

After alignment, make sure to address any errors to maintain the quality of your dataset.

Error Correction Guide

Error correction is a crucial step to ensure your data is accurate and reliable. Follow these steps:

Check Calibration: Even a small calibration error, like 0.1°, can lead to major deviations .
Strip Alignment: Use specialized software to achieve sub-centimeter positional accuracy and sub-degree angular precision .
Monitor External Factors: Keep an eye on GPS signal quality, weather conditions, and IMU (Inertial Measurement Unit) performance .

"Proper calibration is crucial for the successful utilization of the LiDAR payload. The laser scanning system will be unusable without proper boresight calibration and strip alignment." - Inertial Labs

Tools like Anvil Labs can simplify these error correction steps, making it easier to handle large datasets with multiple data types effectively.

Accuracy Measurement Guide

Ground Truth Verification

Ground truth ensures the precision of 3D models by comparing them to real-world measurements. This involves collecting data from flat surfaces like parking lots or tennis courts to serve as a reference. Using high-accuracy GPS surveys on these surfaces can achieve an RMSE (Root Mean Square Error) of ±1.2 inches (±3 cm).

Surface Type	Recommended Collection Method	Typical Accuracy
Flat areas (e.g., parking lots, tennis courts)	GPS survey points with high-accuracy receivers	±1.2 inches (±3 cm) RMSE

When gathering ground truth data, make sure to record:

Date and time of the survey
Surface type or material being measured
Weather conditions at the time of collection
Collection method used
Equipment calibration status

Error Analysis Methods

A comparison between iPhone LiDAR and Leica BLK360 shows that 90% of point clouds deviate by less than 4.8 inches .

For a detailed error analysis, focus on these key areas:

Horizontal Accuracy: Photogrammetry can reach precision levels as fine as 0.4 inches RMS under ideal conditions .
Vertical Accuracy: LiDAR systems typically achieve an absolute vertical accuracy of about 1.2 inches .
Point Cloud Density: Check if the spacing between points meets your project's resolution needs.

Be thorough in documenting these metrics in your accuracy report to ensure clarity and consistency.

Report Format Guide

Organize your accuracy report with the following sections:

Report Section	Required Information
Equipment Details	Specifications of devices used and their calibration status
Collection Parameters	Flight height, overlap percentage, and weather conditions
Ground Control	Number and placement of Ground Control Points (GCPs)
Verification Results	RMSE values, point cloud density, and any coverage gaps
Environmental Factors	Lighting conditions, surface types, and potential obstacles

For larger or more complicated sites, tools like Anvil Labs can help you manage and visualize accuracy data from various capture methods. This makes it easier to pinpoint areas that may need further verification or additional data collection.

Summary and Next Steps

Main Checklist Points

Here’s a breakdown of the key requirements for photogrammetry and LiDAR, based on earlier standards:

Phase	Key Requirements	Accuracy Target
Pre-Survey	Calibrate equipment; set up GCPs	±0.016 feet RMSE
Data Collection	30+ checkpoints, monitor weather	Within project specifications
Processing	Independent source verification	Double dataset accuracy
Quality Control	Error detection, SDASN analysis	±0.4 inches horizontal

These steps create the groundwork for the recommended tools and suggested actions below.

Recommended Tools

Here are some tools to streamline processes and support the checklist:

QA/QC Toolbar: Enables systematic project reviews with tailored issue lists.
Feature Analyst Tool: Helps sort and filter accuracy issues based on specific criteria.
Scale Management: Offers predefined zoom levels to simplify error detection.
Export Capabilities: Allows shapefile generation for stakeholder reviews.

Additionally, Anvil Labs provides centralized 3D model management and data integration to enhance workflows.

Implementation Guide

Using the checklist and tools, follow these steps to ensure precise results:

Equipment Setup and Calibration
Calibrate sensors to maintain LiDAR accuracy between 0.02–0.39 inches. Choose suitable image sensors tailored for photogrammetry projects.
Data Collection Protocol
For example, the TrueView 680 achieved an RMSE(z) of 0.011 feet by employing:
- Automated 3D target detection
- XYZ offset corrections
- Surface normal deviation analysis
Quality Assurance Process

"By opting for a multi-technology survey protocol, we are more flexible in our building coverage strategy and can adapt to the specificities (size, height, accessibility, etc.) of each zone." - MyDigitalBuildings

Use LP360's automated tools to detect outliers, noise, and systematic errors. Export identified issues as shapefiles to facilitate team reviews and track corrections effectively.