Shamrock, Ireland

Royal Docklands, London. Combined mulitbeam / laser survey

Royal Docklands, London. Combined mulitbeam / laser survey

London Eye, London. Laser survey

Before and after cleaning

Qloud is a fast and powerful data validation tool, used for cleaning and validating large data sets.

• Stastistical data cleaning using CUBE algorithm and TPE values
• Automatic Surface Spline cleaning
• Manual cleaning

Technology marches on with the newest echo sounders capable of thousands of soundings per second. Company profit margins dictate that additional time spent in manual data inspection is just not viable. On the contrary: companies look to slash post processing time even as data volumes increase exponentially.

In the face of data deluge, QPS' philosophy of computing almost final soundings in real-time remains as effective as ever. However, as always, online results must be validated using offline tools. But current and future data volumes require new, more powerful offline tools, primed for the future.
Qloud, the latest software offering from QPS, is just such an offline tool that is fully integrated with QINSy, the online acquisition software, but also available as a stand-alone package. Handling extremely large data sets, it performs statistically-based data cleaning using parameters such as Total Propagated Error (TPE), IHO S-44 guidelines, in the CUBE algorithms (UNH), and in the Least Square Statistical Spline Method.

Qloud imports DTM points from QINSy QPD files, (includes multiple data attribute flags generated in QINSy), or from any third party point files, with or without attributes and metadata. The moment the data is loaded into Qloud, the survey is viewed as a single cloud of data points, presented in the full geographical context of ENCs and GeoTIFF imagery.

Automatic data cleaning tools using Clips, Area Spline filters and CUBE, are applied to the entire survey at once, or sequentially to selected sections. Where centre and outer beams overlap, TPE values are used to correctly weight each data point. A full complement of manual editing tools is also available.

Whether the data is viewed as individual soundings, or as gridded data, is the user's choice, often dictated by the deliverable format. For example, viewing every last individual sounding is wasted effort if the deliverable is a single mean sounding per grid cell. Flexible viewing options allow, for example, presentation of the SD attribute of a gridded dataset in one pane, and the 3D points cloud in another pane. By pinpointing bad SD values, the spotlight is quickly directed in the points cloud for more focused analysis.

Validated data can be exported to:

• A QINSy Sounding Grid file where the different CUBE hypotheses are presented, each on a different layer, together with their corresponding statistical parameters such as confidence level and count.
• Dedicated CUBE ASCII files containing grid points and statistical parameters for each hypothesis
• ASCII points file