Survey Grade Accuracy

 When conducting drone surveys, understanding how accurately the mapped data reflects real-world measurements is crucial. One of the most important methods to assess this accuracy is Root Mean Square Error (RMSE). RMSE is a way of measuring the average deviation between the measured points from the drone survey and the true points from ground control or checkpoints.

For land surveyors, RMSE can be thought of in a similar way to how we use least squares adjustments in traditional surveying. In least squares, we aim to minimize the residuals or errors between observed measurements and true values, and RMSE helps us quantify these types of errors in the drone-derived data.

 Several factors contribute to errors in drone surveys, affecting overall accuracy:

• Photogrammetric RMSE – This error represents the accuracy of the generated map or model. It depends on factors such as camera quality, flight altitude, and processing methods. For well-controlled drone surveys, the typical RMSE values are:
  
  • Horizontal RMSE: ~0.08 feet
  • Vertical RMSE: ~0.12 feet

• RTK GPS Survey Error – When RTK GPS is used to set ground control points (GCPs), the precision of these points is typically around ±0.1 feet, both horizontally and vertically.

 Errors from photogrammetry and RTK GPS don’t simply add up directly. To combine them, we use the Root Sum of Squares (RSS) method. This is similar to how, in traditional surveying, least squares adjustments account for multiple sources of error by minimizing the sum of squared residuals.

The formula for total accuracy in a drone survey is:

Total Accuracy = √(Photogrammetry RMSE² + RTK GPS Error²)

Example:

• Photogrammetric vertical RMSE = 0.12 feet
• RTK GPS error = 0.1 feet

Using the formula:

(0.12)² + (0.1)² = 0.0144 + 0.01 = 0.0244, resulting in a total vertical accuracy of approximately 0.16 feet.

Even though the individual errors are small, they combine to increase the total uncertainty in the survey.

 To ensure that your drone survey accuracy stays within 0.1 feet, it’s important to minimize all sources of error as much as possible. Here are some best practices:

1. Use High-Quality Ground Control Points (GCPs): Properly placed and accurately surveyed GCPs help to improve the alignment of your data, reducing overall RMSE. This step is akin to using well-established control in traditional surveying for least squares adjustments.
2. Optimize Flight Settings: Flying at lower altitudes and using high-overlap images improves the accuracy of photogrammetry, much like maintaining precise field measurements in traditional surveys.
3. Calibrate and Validate RTK GPS Equipment: Regularly check your RTK GPS measurements against known benchmarks to ensure that the ground control points are accurate, just as you would check field data during a least squares adjustment.
4. Implement Quality Control Checks: Just as you would verify adjusted survey data, compare your processed drone data with independent checkpoints to confirm that the accuracy meets your required standards.
5. Standardize Data Processing: Using consistent software settings and processing techniques helps to avoid errors that could distort the final deliverables, similar to ensuring proper methodology when performing least squares adjustments in field surveys.

By following these best practices, land surveyors can maintain accuracy within 0.1 feet, ensuring the reliability and precision of drone survey results for clients, while minimizing the combined errors just as we would in traditional surveying techniques.