Accuracy vs. Precision - what's the difference and why they matter?

Published by

Joshua Spires

on

March 21, 2025

Accuracy and precision are often confused in drone surveying, regardless of the payload being used, such as RGB cameras or LiDAR sensors.

Let’s break down what each term means and why they matter.

What is accuracy?

Accuracy refers to how closely the capture data matches known real-world coordinates. In other words, it tells you if your data lines up with the actual position of things on-site.

There are two types:

  • Absolute accuracy - How well your data matches known ground control or GNSS coordinates. It’s about real-world alignment.
  • Relative accuracy - How well objects in your dataset align with each other. Your data might be consistent within itself, even if it’s not perfectly tied to real-world coordinates.

What is precision?

Precision is all about repeatability, how consistent your data is across multiple captures. Even if your dataset isn’t perfectly accurate, if it captures the same thing the same way each time, it's considered precise. This matters when you're surveying the same area repeatedly and need to compare data over time.

A top down (2D) view of a stockpile.

Why do they matter?

Accuracy and precision influence how confident you can be in your data, and the decisions based on it.

  • Accuracy ensures your maps, measurements, and models reflect the real world.
  • Precision ensures you can compare data across different surveys and still trust the results.

Example: Measuring a Stockpile

Let’s say you’re using a drone to capture the size and location of a stockpile on your site.

  • If your data is accurate, the stockpile in your model aligns perfectly with its actual location on the site, every edge, corner, and volume reading matches the real-world coordinates.
  • If your data is precise, repeated flights over the stockpile produce consistent results, the stockpile shows up in the same place, with the same shape and size, every time.

Here’s where it gets interesting:

  • A dataset can be precise but not accurate, meaning the stockpile appears consistently, but in the wrong spot.
  • Or accurate but not precise, meaning it aligns with the real world, but changes slightly with each capture.

The goal is to achieve both, high accuracy and high precision, so you can reliably track changes, report volumes, and make confident operational decisions.