Tag Archives: 3-D

Overlapping Photogrammetric systems

View of Point Cloud from side/rear

Spending a lot of time seeing how different photogrammetric systems and point cloud software packages speak to one another. sometimes the interface is elegant, sometime not so much…

Advertisements

Aerial Photography > Dense Point Clouds

Since 2009 I’ve done a fair amount of work with aerial photography, all the while continuing to explore architectural photography and photogrammetry while on the ground. I’ve tried in the past with some success to merge the two things but have found the results to be lacking something. That something, I think, is the hyper dense point cloud – and now I have it pretty much figured out.

Below are tow animated GIF images showing first the point cloud describing our local hospital (which I was contracted to cover with aerial photographs a few years back) and the second is a textured 3-D model of the scan.

Its so amazing to think that the flight over this structure can be reconstructed virtually a few years later with even more options to “fly around”…

Dense Point Cloud

Dense Point Cloud

Textured 3-D model

Textured 3-D model

Point Cloud from Aerial Photos using photogrammetry

Here is a shot of the quarry in Staunton from the air, shot back in May:

Quarry

This is only one of a series of photos that I processed into a large 3-D point cloud that captures the real world shape and size of this big hole in the ground. The point cloud can be imported into AutoCad as tens or hundreds of thousands of discreet point measurements (along with true color values in RGB). This approach is very cumbersome in AutoCad but can be useful to then creat a 3-D mesh and eventually a “solid” in order to make difficult volume calculations.

Or it can be exported to a text file/table and brought into AutoCad (11 or later) as an indexed point file (*.pcg) which is much more easily managed in AutoCad.

Here is an animated GIF file showing the three dimensional nature of the point cloud.

Quarry point cloud viewed in AutoCad

Page Updates: Point clouds, F&M, Roxboro

I’ve updated my page on point clouds with some new images from Franklin and Marshall’s Goethean Hall that show some of the true color point clouds generated though photogrammetry. Likewise, I’ve update the page on the Roxboro House project to include some point cloud views.

Reclining Figure Arch Leg – Sculpture by Henry Moore 1973

Reclining Figure Arch Leg – by Henry Moore 1973

This summer (2010) marks the real debut of a reliable automatic generation of dense point clouds by using photogrammetry.  I travelled to Switzerland in July and gave this exciting new software module a try on something that would be a real challenge to document in anyway other way – a sculpture by Henry Moore.

The four panels below show different views of the point cloud generated from these photographs. For me the real lesson in this study is that objects to be documented in this way need to be photographed in conditions that produce as even lighting as possible. One can see that some of the reflectiveness from the metal (brass) surface of the sculpture resulted in less-dense-than-desired regions in the point cloud. Still, I was excited to see how well it performed under what I now understand to be relatively harsh conditions.

The density of these point clouds is considerably higher than what photogrammetry ordinarily would produce while not being the hard-to-manage overkill produced by laser scans. what’s more, these point clouds can be enriched with the same type of smart point clouds that provide a meaningful link to both rectified photography and architectural description.

Andrea Pozzo

2D Descriptions of 3D conditions

2D Descriptions of 3D conditions

What I’m talking about…

Andrea Pozzo says it all with this great drawing.

The Roots of Photogrammetry

 

Stereoscopic View of Worlds Fair Paris 1900

Stereoscopic View of Worlds Fair Paris 1900

The roots of photogrammetry lie in sterescopy, capturing three dimensional views with pairs of images.  This is almost as old as photography itself.  This image I particularly like because it looks like two capital letters “A” – like my last name – and it is in Paris, where I first learned about photogrammetry.

Franklin & Marshall Exterior Survey

Old Main at Franklin and Marshall

Old Main at Franklin and Marshall

Last spring I did some work at Franklin and Marshall College in Lancaster, PA.  I prepared a combination of measured drawings in CAD linked  to rectified photographs of the three oldest structures on the campus. These were then used as base documentation for an assessment of the existing conditions of the masonry surfaces across the extent of these buildings’ exterior facades.

The sequence of this work went roughly as follows:  I first shot photos of all of the buildings from points of view available on the ground (using a specially calibrated digital SLR).  Back in my offices in Virginia I used photogrammetric software to “back calculate” the camera stations and to make precise 3-D measurements of points on the three structures.  I used this dimensionally accurate point cloud as a reference to create rectified photographs of the surfaces visible from grade.

F and M Old Main Rear Elevation

F and M Old Main Rear Elevation

Using the point cloud and the rectified photographs I then created measured line drawings of the structures and laid them out onto tabloid size sheets for use in the field.  These were taken by hand up into an aerial lift so that conditions could be noted with a fair degree of accuracy once assessment were made both visually and manually.

tabloid size field sheet with notes

tabloid size field sheet with notes

Once the assessment was completed the data recorded on the field sheets was entered in the CAD drawings.  At this point the line drawings in CAD were enriched by a mosaic of rectified photographs visible inside of AutoCAD.  This allows for an accurate transfer of notes from the field to eventual construction documents.  A sketch of an area or region can be transferred to CAD and become dimensionally reliable.  This is a very important step because it provides for an accurate tally of areas to be treated in one way or another.

aerial lift

aerial lift

Since the main structure was rather tall, I also returned to the site to go up in the aerial lift to collect more photographs to further enrich my drawings with better rectified images of areas that were either blocked by vegetation, neighboring buildings, or were too foreshortened to provide good rectified images.

F and M Old Main Rear Elevation with photos

F and M Old Main Rear Elevation with photos

Discovering Photogrammetry in France

I first encountered the concept of architectural photogrammetry while working as an architect in Paris in 1990 -and was immediately intrigued. So much so that I decided to swap my job with a 45 minute commute into the city for a new job just a short bicycle ride away (in the town of Egly)  in order to learn how to do this work.  Simultaneously I had to learn how to draw with a computer; up until then I had only drafted with pen and pencil (plastic leads for working drawings).  It was a lot to take in, but I think its fair to say that I loved it.

digitizing puck

digitizing puck

Back then, the process was nearly completely analog: We shot film and had it developed into 8″x10″ prints.  These were then taped onto a large format digitizer so that points on the photos could be communicated to software running on a PC.  Then after a photogrammetric analysis of this information was completed, we made drawings that were plotted to scale – and it was these prints that were the “deliverable” product desired by our clients.

I remember seeing a number of photos taped onto a digitizing table and watching my colleague Herve use the digitizing puck to point to the same location on a building as seen from various points of view and then push a button and seeing the point location appear in 3-D coordinates in an CAD drawing on a computer screen.  It seemed magical.

As more and more points were measured in this way, I could start to see the shape of the building.  This became even clearer as points were connected into lines and polylines; the “point cloud” was becoming a “wire frame”.

When I first got out of architecture school and was working for a firm in New York City, the first thing I had to do was to learn how to make measured drawings of existing conditions.  It was messy, confusing work on site and often yielded error-ridden results in the office that required repeated return visits to understand and clarify what was really happening architecturally.  When I saw a dimensionally accurate 3-D model representing an existing structure being created from a handful of photos, I was hooked.  Nobody I knew was doing anything like this in the United States.

 

Triumphal Arch Jardin de Tuileries

Triumphal Arch Jardin de Tuileries