Case Study: Optech

Optech LMS: Streamlining the Lidar Calibration & Data Processing Workflow
By James Wilder Young

Data calibration is critical to lidar processing. Because the ability to optimize and maintain a sensor calibration is essential to reliable and consistent data quality, the new Optech LMS software, which automatically optimizes sensor calibrations by using available project data, is intriguing.

This case study evaluates Optech LMS’s ability to improve output lidar data quality, and identifies software features that improve the workflow efficiency.

Lidar system calibration requires validating and adjusting sensor parameters—roll, pitch, heading, and elevation bias, and sometimes factors such as scan scale.

Optech LMS utilizes redundant information associated with flight line overlap areas to derive a dynamic sensor calibration model that is applicable throughout the entire project. Incorporating pre-determined calibration statistics, computed XYZI outputs are generated in a variety of industry standard outputs.

There are two calibrations, installation (initial) and mission (subsequent), with different but equally important configuration requirements. Installation calibration occurs whenever a lidar system is installed in an aircraft, and mission calibration whenever the aircraft flies to collect data.

Mission calibration requires one cross-line perpendicular to parallel flight lines, with two cross-lines preferable.

In the ALTM workflow, a raw lidar file (range, angle, timing, attitude information) is first decoded to enable inertial processing (Figure 1). Optech LMS requires the input of an inertial solution, raw lidar sensor data, and sensor calibration/configuration files. These are combined to compute XYZ coordinates.



Optech LMS has six processes: Project Setup, Block Identification, Planar Surface Extraction, Lidar Rectification (via bundle adjustment), Results Analysis and Data Output.

Project Setup and Block Identification use information from the Flight Management Software flight plan to ensure that flown flight lines match planned lines.

The automated Planar Surface Extraction algorithm searches for surface planes within the data. Redundant planes are identified in overlap areas and thinned for efficiency.

Analysis tools then compare planar surfaces to ground control points/surfaces to assess whether Lidar Rectification is warranted. If rectification is preferred, its speed is optimized with distributed and parallel processing for simultaneous processing of missions and tasks across a network.

Optech LMS applies sensor corrections derived from adjustments to the sensor calibration file. Parameters are derived globally for the project or by flight line.

Block adjustment (rectification) has four steps: roof line comparison, tie plane determination, tie plane selection, and self-calibration. Although parameters can be extensively tailored, default settings work well.

Optech LMS outputs LAS files in 1.1, 1.2, and 1.3 formats, as well as an ASCII output. Users can select input and output reference frames using embedded Blue Marble Geographics toolsets.

Methodology

The case study involved a sensor installation calibration based on a series of flight lines flown at a given altitude.Two flight lines with 50% sidelap were flown in parallel, with one perpendicular calibration line.

GPS base station data was acquired both from CORS and from data collected independently on National Geodetic Survey (NGS) points. The inertial solutions yielded excellent results, with one poorer solution identified and resolved by Optech LMS.

Although this case study used only lidar data, ground control or survey data can also be added within Optech LMS to further refine the results. Optech LMS also aids in resolving GPS biases to provide excellent strip-to-strip coherence.

The accuracy analysis of all tie planes from standard processing and after rectification is shown in Figure 2, and demonstrates excellent agreement.

Conclusion

Optech LMS significantly streamlined calibration and LAS output. Once the mission and project were created and configured, the automated processes ran on machine time. The interface is easy to use and the ability to customize the calibration process is exceptional. We now expect to spend less time calibrating and more time generating value-added product.

---

James Wilder Young is currently the Lidar Solutions Specialist at Aerometric, Inc.

» Back to our Aerial Mapping 2011 Issue