This lecture was presented at the 3D Digital Documentation Summit held July 10-12, 2012 at the Presidio, San Francisco, CA
High Resolution Digital Photogrammetry with Object Surface Texture
Digital Photogrammetry when coupled with multi-image capture can provide high resolution
imagery with detailed surface texture. Surface texture can be in greater detail than traditional
point cloud laser scanning, and the camera is not required to be in a fixed location allowing
greater flexibility in capturing raw data.
Existing building documentation or other large objects are example cases where it can be difficult
or impossible to document due to lack of access. Often owner and architects resort to repelling or
other means of surveying an object. Digital Photogrammetry can be obtained remotely obtained
controlling a camera as it moves across an object while the camera and rigging supported by the
object itself e.g. swing lift, without the need to put human in place.
We will discuss the use of robotic equipment originally developed for aircraft skin diagnostics
has been modified to allow for use at an appropriate distance from a façade to capture detailed
imagery. Large composite images are created from the detailed close-range images. These
comprehensive composites allow for wide-area pan and zoom inspection of the façade by all
project stakeholders viewing safely online as if they were right there.
Presenting high quality information in a usable format is the key to guiding a successful project.
In this session we will describe how 21st century photogrammetry coupled with web based
technology was used to document and guide an 1896 church masonry restoration project. This
design delivery method resulted in more project control, reduced equipment and manpower effort
during the investigation phase, and will provide the next generation of preservationists and
owners a readily accessible record on which to base future projects.
How photogrammetric imagery was used to produce blueprint substitutes with full surface stone
texture will the demonstrated in this presentation. These blueprint substitutes enhance
productivity, development of work scope, and project management. High resolution photo
inspections viewed interactively side-by-side online with these blueprint substitutes were used to
assist in RFP development and bidder review of the project.
We will demonstrate how the documentation tools allowed for remote, close-up assessment of the
stone and other surface features on a scaled rendering. The web based display facilitates
interactive use, and “on-image” recording of forensic investigation data e.g. drill resistance
testing, ground penetrating radar, impact echo and ultrasonic graphs and data, for before and after
restoration comparisons. How the technology makes work scope and forensic data retrievable,
accessible, and understandable to the field user will be illustrated. Photogrammetric scaled
photo-renderings with enhanced surface texture as blueprint substitutes and high resolution photo
inspection tool allowing pan and zoom capability will be demonstrated to show how greater
control can be established for projects, and reduce bidder uncertainty.
These tools ultimately allow for the evaluation and documentation of the conditions of stone and
other object surfaces not possible until current development in computing and the internet.
Brown: Thank you Mary. I’m really approaching this from a practitioner’s point of view. I was the property chairman of a historic building asset for a decade, and we were marching towards a masonry restoration project for that building over most of that decade. The techniques that you’re going to see today were developed to make the information and data developed by conservators and engineers and others more available and accessible by the team of people that were going to work on the building and to create a more permanent record of what was done.
So we wrapped up yesterday’s session with the question, “is it science or is it practical and useful” and I think the answer is yes or at least from my perspective. There’s a lot of issues that we face, and I know that how many of you have faced some of these issues in the projects that you’ve managed. You know solving some of these things require some new tools or either old tools to be used in new ways and a more integrated approach to the work at hand. Solving them will lead to better and more efficient work in the short term and a better record for the next generation of owners, conservators, preservationists, and historians.
So how do we take a complex 3-D building like this one hundred and twenty year old William Halsey Wood designed Church of the Ascension in Pittsburgh and translate it into the visual communication tools, namely blueprints, that the architecture, engineering, and construction industry is familiar with. You know on this particular project, this was a building that I was the property manager for, we had a wonderful forty page conservator’s report on this building but the information was never going to be used. It was effectively inaccessible due to a text based format. We needed to convert the data and the report to more useful information and instructions. For example, how do we locate the scope of work or the quantity of linear feet of mortar repointing on a project like this one? How do we provide a more accurate and higher quality information as part of our RFP provided to potential bidders? You can see here the building is quite complex, a lot of three dimensionality to it.
So this is the current state of affairs or was the current state of affairs, hand measured drawings, which may or may not be that accurate. Measured drawings are conceptual in nature. They provide little additional information compared to the original construction drawings and so I was looking for a solution for this building. Wouldn’t you have liked to have had a blueprint like this with all of its surface texture for that project. This particular image, these mortar joints that you can see here are in the range of a half inch to an inch in width. The original images were produced at the four to eight pixel per inch resolution range, which was good enough to produce a 1/24th scale arch-D 36 inch wide sheet to include in the RFP process for this project in 2008.
So let’s take a little closer look at the process here. So there is the original photography being rectified and we can zoom in and begin to see some other aspects of the building. Then we do the conversion process to the blueprint substitute, we scale it, we add scope, we put it in the page makeup, and I’ll run that one more time for you so you can see what’s going on there.
I’m going to show you some other examples now of different buildings that we’ve done different changes. So this is an arch-D page and you can see the facade of this robber baron era house in the Pittsburgh region. We’ve got quite a bit of information here. You can already see the water plant growth staining here in the upper portions of the building even at small scale. And as we zoom in a little bit, you can begin to see some of the tooling of the stone. The surface textures are different and if we come in here closer, you can actually read the historic landmark plaque and see the different textures. If we jump up to this crest here above the door, you can see that if we had to reproduce this crest in the case of a natural disaster or loss, we’ve got pretty good information just from this photograph. When we jump up to the upper area, you can see the mortar joints are a lot wider up here at the top, so you’ve got some thermal movement or something else that’s happened through the years there in that building. So that’s just an example of a little different scale of resolution that is available.
So here is a terra cotta example. This is the Mellon Institute in Pittsburgh which is owned by Carnegie Mellon. It’s either, I don’t know if the color is really showing up well here, but we’ve got a lot of water staining and green algae here in the corner under this piping. But as we zoom in to the terra cotta there, you can see that we were able to extract the full pattern on that terra cotta into the blueprint substitute and makes it very useful for marking up the drawings, determine scope of work, and then, sorry I have to come back over to the other side here. We can accentuate different features of the building depending on how we process the information. So you can see on the left, we have more of the mortar joints and on the right we have more of the stone texture. We can even get a quad view here so we have color and a variety of different things here and we’re all moving in comparison. So with this kind of thing you can do a preexisting conditions or post demolition, post repair, and you could even monitor the project over time. You can see very different features brought to bear there.
So this is a set of records from the HABS collection at the Library of Congress. You might recognize this as the Alamo and we’ve got a hand measured drawing with an actual photo of the facility here. So we don’t have to go create new data to use some of these tools. Sometimes we can use historic data and come up with good information. However, we may want to create new information depending on what’s going on. These two photos are taken seventy years apart and if we look at the capital here on this column, you can see we have a great deal of material loss here in this capital and we have continued deterioration in this spot here as compared to the location seventy years previously. If you look at this location here, you can see that we have some damage that existed seventy years ago.
We can do very high resolution. This is gigapixel photogrametry here. These are multi image compilations and that’s about at the resolution that we were before on this building. If we come on in here, you can really get right in there and do your visual inspection. You can actually see where they put aluminum flashing here on top of this capital and then it’s forged on with some cementitious material here in this place where the edge is not clean and then we come over and we have the flashing again. So you can have that kind of very high resolution inspection available to your RFP process and what better to have available to them.
Another particular issue that we deal with often is difficult to access buildings. I’m about two floors below grade here looking up this hill at an eight to nine story clock tower at the Jefferson County Courthouse in Port Townsend, Washington, and we were able to come in here and document this clock tower and you can really zoom right in there, you can see some loss of mortar here and then you can see that there’s a chip missing from this stone here. So we can do fairly high range inspection even from difficult to access material. So this should allow your conservator to do some of his inspection from his desk and reduce the field effort to validation.
Everything I’ve shown you so far, the camera has been fixed in a single point of view. This particular image is what I would call a linear panorama where the camera is marching along a parallel line to the surface of this frieze. This is probably about fifteen columns by about three or four rows of information here and we can jump in here and we’ve got pretty good detail here. This was not intended to be a super high resolution but if we go along to the other end, you can see we’ve got good detail here in this frieze.
So that’s the core tools. Then going a step beyond that, we can then add annotation capability to the system so we can annotate on image. I’ll just turn those annotations off for a minute and we can go to different areas and we can jump right in there. If I turn these annotations back on, you can see here when I mouse over the grates on this building, you’ll notice and it may be a little difficult to see, the original wooden muntins over the glass here are still available above the door and then down here in the lower area, we have a new aluminum forceable entry post 9/11 upgrade to the building. This is the Hiram Chittenden Locks in the Seattle area and so we have a very odd shaped annotation here where we were able to annotate both areas at the same time. We can also put forensic information on the building and I’ll show you that in another example.
Again, we have these points of interest that we’ve created here and we may have different elements. Here we have a moisture study that was done at this location. We’ve color coded the data points for their information and we can get a pop up, a message describing the date and the measurement and if we click on some of these points, we can drill down to the engineers report, who took the data test. Obviously we can link to any other kind of data and if you don’t like that presentation, we can present it in a graphical format or this could be a repair specification or almost any other kind of information that you can basically imbed on image.
What we’re trying to do here, the solution was to try to come up with a suite of tools that could solve problems in the construction industry for managing and controlling work. If you develop the blueprint substitutes and then you have photometric gigapixel surveys, obviously then you can put your work scope, you can have a web discussion, you can bring in outside experts and even oversees folks. Through the use of web technology you develop your scope and specification and move those back onto your blueprint substitutes and then that can provide greater accessibility, better data into the RFP process and of course it can provide the historic record. It also has several other benefits in terms of clients, websites, public accessibility to features and aspects of the building that might not be viewable from ground level or you can also generate interest in your project, fund raising perhaps, membership, things like that. That’s how we’ve seen the tools used in the last four years.
The benefits to you as conservators and technicians and owners are tremendous. We found that we can reduce equipment and manpower in the field to validation of existing conditions rather than discovery so the conservator may spend as much time on visual inspection at his desk but we’ve reduced the need for cherry pickers and repelling and other kinds of things in the field to more of a minimum to a validation effort. We provide much greater information to the perspective bidders and other people that are interested in working on a project. Risk premiums are reduced and therefore bidding tends to come in a little bit tighter. So that’s what we’ve got. These are basically gigapixel photogrametry with a conversion to blueprint substitutes so that you’ve got all that surface texture and then we couple that with these other tools.
So, I think that’s really all I have to say about what you might do with it today but then the advantage of course, is that you have better records that you can leave to the next generation and if you need archival material, you can always output the drawing pages to velum with some carbon based inks on your inkjet. Theoretically, hopefully, we’re solving current problems and leaving the scientific record for the next generation. If you want to trust digital, well you can, but I don’t recommend it.
Battle Brown created the patent-pending PhotoDrawing™ image rendering and PhotoSurvey
inspection tool process to provide detailed high surface texture scaled drawings to the
architecture, engineering, and construction industry for existing buildings, structures, and objects.
He has over 35 years experience in imaging, graphics, and communications, software and has
been involved in the AEC industry for over a decade as owner, contractor, and consultant. He is
a Senior Research Scientist at Carnegie Mellon in addition to being a serial entrepreneur.