Scanning and Drawings

Two-dimensional Architectural Drawings

Base plans were first created in campaign one and expanded upon in campaign two, providing Pecos with an extensive record of extant architectural features.

First, the 2021 team, working with all site's documents, created a single accurate base plan that contains the largest amount of site-based data that currently exists. The value of the base plan is that all the data is correctly placed, spatially, relative to each other. Within the compiled base plan, newly created wall segment number conventions could be overlayed onto line drawings. The new wall segments allow for more specific identification of individual masonry units, improving the quality of documentation and recording on site. The base plan is a living document meaning over time as more accurate data is created, it can be done so either directly in the base plan drawing, or through the use of external referencing (an option available in AutoCAD), to ensure that everything properly aligns.

The end result is that staff do not have to search for a plan that provides the necessary data for their project because all pre-existing data is now held within the layers of this single base resource.

Convento Plan with Room IDs Overlayed

18th Century Church Plan with Point Cloud

Point Cloud Capture

In conjunction with the creation of the base plan, a Point Cloud was used as an alternative method of spatial data collection in campaign two. A point cloud is a set of points in three-dimensional space that represent individual locations on the surface of an object or objects. Each point is assigned coordinates on the X, Y, and Z axes. This allows the points to be viewed together as a three-dimensional unit composed of points in a “cloud” rather than with solid surfaces (see examples of a point cloud below).

Point clouds and laser scanning technology are powerful documentary tools for historic preservation and have been employed successfully in the recordation of historic buildings, archaeological sites, and cultural landscapes. The ability to digitally capture building and landscape features with high levels of precision and accuracy and a user-friendly device allows preservationists to document elements efficiently without compromising resolution. For this reason, the Pecos HSR project crew utilized laser scanning to generate architectural basemaps of Pecos Pueblo’s northern and southern units and the site’s Defensive Wall.

Through experimentation, the team developed several techniques by which to improve registration quality and decrease the error rate:

Using closely spaced two-minute scans over fifteen-minute scans with farther spacing increased the registration success rate.
Positioning fifteen-minute scans at high points in the site and connecting these with shorter intermediary scans decreased the error rate and allowed for more detail to be captured in plan view.
As the site’s topography is variable, scanning on high points also ensured better visibility of the limited smooth surfaces with which Trimble Perspective could register.
Organizing scans into radiating starburst patterns limited accumulated error that would be expected to occur in long linear arrangements (shown in the photo below)
A “carpet-bombing” approach to filling in gaps in point cloud data proved to be more time-effective, as registration could be finessed outside of the field.
Using trilateration to position a station closer to its accurate location could be achieved by repeatedly breaking and relinking the station to multiple other scans in its vicinity, allowing the station to move itself into place.

The laser scanning data collected during this campaign was compared with drone photography to determine which method of data collection could be the most useful for future projects.

Field Crew positioning the scanner in the field

Example of the “starburst” pattern used to minimize accumulated error in registration between scans from Trimble Perspective software during 2024 PECO fieldwork

resulting point cloud of the above laser scan

An Analysis of Best Methods for Producing 2D Drawings

Lastly, architectural drawing sheets for the pueblos and defensive wall were produced from the point cloud data in combination with both Alfred Kidder’s excavation maps and the topography data created by Archaeology Southwest in 2015. The first drawing of the extant walls of the South Pueblo, produced from the 2024-point cloud data, illustrated the numbering schemes employed at the site and delineated the wall segmentation to be used in the upcoming conditions assessments. The second drawing sheet included the architecture of the North Pueblo, demonstrating the alignment between the extant wall segments documented in the point cloud, the buried architecture recorded in Kidder’s excavation map, and the topographic data collected in 2015 by Archaeology Southwest.

These drawings were produced using a collection of three softwares: AutoDesk Recap, CloudCompare, and AutoCAD, processing over 500 gigabytes of raw data into digestible drawings for future use. This resulted in a multi-step process:

First, to reduce file size much of the surrounding extraneous information was deleted, leaving only the architectural data in place.

Second, the decimation tool in AutoDesk Recap was used to reduce the number of points in the point cloud while maintaining object definition.

Third, the point cloud was segmented into two separate point clouds to further reduce file sizes: one containing the South pueblo and the other containing the North Pueblo and Defensive wall. After drawings were created, these files could be re-registered together into one main file for future use.

While it was possible to generate the two-dimensional drawings from this point cloud, the quality of the output is heavily dependent on the resolution and continuity of the point clouds.

The results of this analysis indicate that a high-resolution aerial drone produced topographic lines that were far cleaner and more readable than any produced by a terrestrial laser scanner, regardless of the processing software. Further, the drone is cheaper and more time effective than the scanner. Thus, consideration should be taken when employing laser scanning in future projects at the site.

Page updated

Google Sites

Report abuse