Recently in Data Acquisition Category

Mark Willis comes up with another awesome application of photogrammetry to rock art studies.

Follow this link to the entry on his blog.

This post illustrates the use of a simple KAP rig in conjunction with structure from motion photogrammetry to generate a dense surface model of a threatened archaeological site.

This post displays a 3D model of one of the sunken plazas from the site of Pukara, Puno, Peru. During July, 2009 I was visiting excavations by my colleague Liz Klarich (Smith) and using the opportunity to practice photogrammetry on large architecture that had been previously reconstructed. The sunken plaza shown was excavated and reconstructed in the 1970's under Plan COPESCO. In the past, I had collaborated with Liz on aspects of mapping the site of Pukara. The sunken plaza has always captured my interest. This model was the first large architectural feature I attempted to record using photogrammetry. The uneven surface of the ichu grass presented some challenges, but these were overcome by aggressive decimation of the surface. The model was generated with PhotoScan and is displayed on this website as a PDF. The file is large, it may take a moment to download, and may strain computers with low RAM.

This post displays a 3D model of a pithouse at the site of Jiskairumoko, Rio Ilave, Peru. The excavations were performed in 2000. The model was generated with PhotoScan and is displayed on this website as a PDF. The file is large, it may take a moment to download, and may strain computers with low RAM. The model was generated as an experiment to see if 3D reconstructions can be created from archival excavation photos. This proof of concept demonstrates that such reconstructions are possible. 

Over at the PublicLaboratory, they have come up with a nice simple and inexpensive telemetry kit for balloon aerial photography (BAP). The information the unit collects should help with image georeferencing. I suspect that the data could potentially be exploited by applications like Palentier that was written by Mark Willis and M. Stange.

Archaeologist and elite low elevation aerial photographer Mark Willis recently posted a wonderful and informative tutorial on how to create a DEM from a Photosynth Point cloud. The photographs used in his tutorial example were captured by KAP, and the site is Palmitopamba in Ecuador. KAP is probably the most inexpensive LEAP technique, so this was a great proof of concept. Mark also strove to use free and open source applications through the entire work flow. Again, this low cost approach means that the methods should be available to nearly any researcher. One of the stages in Mark's procedures draws inspiration from an earlier post on this site regarding the transformation of structure from motion point clouds to real world coordinates. Mark simplified and clarified the process. This was exciting for me because it was Mark's earlier work with photosynth that stimulated my interests in KAP and structure from motion photogrammetry.

This post illustrates low elevation aerial photographic (LEAP) equipment that I use and how this equipment is packed for portability and safety.

ARC 3D, PhotoSynth, and Bundler are all programs that reconstruct 3D scenes from a group of relatively unstructured photographs. All three are based on structure from motion algorithms. However, in my experience there are some important differences in the kinds of photographs that each program is suited to process. This post describes some of those differences.

This post briefly describes Bundler, the structure from motion image processing tool written by Noah Snavely, and details the steps that I took to get Bundler installed on my laptop computer. Related posts include:

• Photosynth to 3D conversion: An Introduction
• Tips: Taking Photographs for PhotoSynth and other Structure from Motion Processors

Structure from motion technologies like PhotoSynth and Bundler are powerful approaches to rapidly generating point clouds. If you are not familar with them, I I posted a brief introduction to the technologies. The image processing techniques produce 3D point clouds. The methods can be implemented on small objects, large ones, or a combination of both. My uses for the technology focus on recording large archaeological sites and their topographic surroundings.

In this post, I'm going to discuss: PhotoSynth, the underlying structure-from-motion technologies (i.e. Bundler), and then describe some of the research potentials that I see arising from the two. I am starting to apply Photosynth and other structure from motion technologies to topographic and archaeological mapping. Related posts include:

• practices for taking the photographs
• strategies for data processing (in preparation).

For now, I'm going to keep this introduction general but take a look at this video that was posted to YouTube. It is a nice proof of concept demonstration of the technologies I discuss.

I have started using PhotoSynth to generate dense point clouds to map archaeological sites. This post contains the RSS feed of the synths that I am creating. A future post will document the processes that I am using to generate the point clouds.

Here is the thread that I posted.

In archaeology it is often useful to represent a surface as a digital elevation model (DEM). With just about any data collection process, there is a tradeoff between the extent of the coverage and the amount of detail that is captured within this coverage. Greater detail takes more time to collect, and smaller spaces can be characterized.

Fieldworkers create DEMs by means of a total station, and this is frequently an efficient method of generating data on surface undulations. However, even with a fast total station it can be extremely time consuming to generate high resolution surface models. Laser scanners may be a useful solution, but these instruments are expensive. Stereo pair photographs can provide a lower cost solution to generating high resolution surface models.

In this post, I describe a workflow that involves processing stereo photographs of a surface to generate a high resolution (50mm) digital elevation model of a space that is roughly 2x4m. The procedure entails generating a dense surface model, or point cloud, in PhotoModeler Scanner. This point cloud is saved as a text file, converted to an Excel spreadsheet, and imported into ArcMap where a DEM is generated.

A new title from the National Academy Press

"As digital technologies are expanding the power and reach of research, they are also raising complex issues. These include complications in ensuring the validity of research data; standards that do not keep pace with the high rate of innovation; restrictions on data sharing that reduce the ability of researchers to verify results and build on previous research; and huge increases in the amount of data being generated, creating severe challenges in preserving that data for long-term use.

Ensuring the Integrity, Accessibility, and Stewardship of Research Data in the Digital Age examines the consequences of the changes affecting research data with respect to three issues - integrity, accessibility, and stewardship-and finds a need for a new approach to the design and the management of research projects. The report recommends that all researchers receive appropriate training in the management of research data, and calls on researchers to make all research data, methods, and other information underlying results publicly accessible in a timely manner. The book also sees the stewardship of research data as a critical long-term task for the research enterprise and its stakeholders. Individual researchers, research institutions, research sponsors, professional societies, and journals involved in scientific, engineering, and medical research will find this book an essential guide to the principles affecting research data in the digital age."

I wanted a device that would serve two simultaneous purposes 1) track caloric effort as I move across space and 2) geocode photographs. Neither of these purposes requires high accuracy GPS, but these tasks do require a small GPS device. The Garmin Forerunner 405CX seems like a perfect device.

This Monday, a new global DEM based on ASTER was released. Through the LPDACC, Craig has obtained a large portion of western South America. These DEMs are now included in the Department of Anthropology GIS data archive.

1. Download data from WIST via LPDACC
2. Wait for confirmation email
3. Pull files from FTP directory indicated in confirmation email
4. Unzip the compressed archive using 7-Zip
5. Move the individual zipped tiles to the root directory and unzip all tile components to this root, a series of tif and xml files are produced. Write over the readme.pdf file, these are identical in all cases.
6. Launch ArcCatalog and build pyramids using the "Batch Build Pyramids" tool.
7. Do something else while that grinds away.
8. More soon...

From the first looks, the data are very noisy. So noisy, they they are not yet useful for archaeological/anthropological research. Perhaps areas where cloud cover is not a constant problem will fare better, but the eastern cordillera of the Andes is not looking good.

The following are a couple of GPS units that are used with a camera for geotagging. I do not use one of these units, I use a Garmin Forerunner 405cx. The advantages are that the unit is a wrist watch and it also has a heart rate monitor. The disadvantage is that the 405cx will only run for eight hours. During fieldwork I frequently have battery problems. Another disadvantage is that the 405cx does not output digital compass information.

GlobalSat BT-359 review

Qstarz BT Q1300 review

ATP PhotoFinder Mini review and another and yet one more and still another.