Laser Scanning Versus Photogrammetry

At the end of my previous post I asked which is more applicable for 3D cultural heritage projects, photogrammetry or laser scanning?

HDS6000 Laser Scanner

A reader pointed me in the direction of the December issue of Geoinformatics, which features an article on page 50 called "3D Laser Scanning and its 2D partners". I wanted to highlight the article in this post as it offers some interesting thoughts on the subject. In particular the article makes the following points (with my own observations):

• There are similarities between photogrammetry and laser scanning. For example, both technologies are used to capture point clouds where points have XYZ coordinates. I would add that the differentiator is that in photogrammetry we are usually capturing points to model a surface (e.g. a TIN or grid) as opposed to a true 3D point cloud (e.g. multiple points with the same XY but different Z's).
• Challenges in the adoption and acceptance of laser scanning, being the (much) more recent technology. For example, the cost and learning curve. I agree with this, however as the authors note this is changing. We face the same challenge in photogrammetry, which still carries a bit of a stigma as a dark art within the broader geospatial community. However times are changing and new technology will continue to flatten out the learning and cost curves...
  
• The all-too-common belief that the two technologies compete. The authors argue that this is a misconception and go on to outline why photogrammetry and laser scanning are complementary. I completely agree with this point: both technologies have advantage and should be implemented as needed on a case-by-case basis. I think we're seeing this in the context of the airborne mapping world as well - an increasing number of organizations are opting for optical and LIDAR systems for simultaneous collection. The Leica RCD105 Digital Camera is a good example of this, as it is typically sold alongside an ALS LIDAR system. There are a lot of advantages for such a system but that's a story for another post.
  

The article then proceeds to discuss laser scanning in the context of recording historic monuments and landscapes, along with several projects (including Cyark) and examples. Overall it is a compelling read and I'd recommend picking up a copy or checking out the online version if you are interested in this topic.

So in summary, I suppose the question above needs to be turned around. For recording cultural heritage, choose the tool set that best fit the requirements - which may mean integrating several technologies.

Angkor Wat in 3D Revisited

Several months ago I wrote about a 3D feature extraction project based on aerial photography at the temple complex of Angkor Wat in Cambodia. While photogrammetric reconstruction is one avenue of documenting historical sites, another method is terrestrial laser scanning. While Cyark has been mentioned before, I thought I'd highlight the foundation again here since it is such an excellent resource for terrestrial laser scanning in the context of archeology and documenting cultural heritage.

Aside from the slick presentation, one of the really nice features of the website is a 3D point cloud viewer, which allows you to navigate various scenes in 3D. While the point clouds are pre-cooked for the viewer (which has a 2 million point limitation), the density is still enough to provide a very realistic experience. You can even see a couple of people in the "Outer Cruciform Courtyard at Banteay Kdei". The process for digital heritage preservation is outlined quite well in this paper.

Angkor Wat 3D Model

Point Cloud: Outer Cruciform Courtyard at Banteay Kdei

One question this raises is which technology, photogrammetry or laser scanning, is the most effective (cost, quality, processing time, etcetera) for cultural heritage projects. For a study on that, check out this paper, which compares terrestrial laser scanning with "terrestrial photogrammetry" (photos are taken from the ground with an SLR camera, not aerial photogrammetry although the processing principles are similar). As one might expect, the study indicates several pros and cons of each method, as well as a look into combining methods. The cost of hardware is higher with terrestrial laser scanning, and the processing (automatic and manual) for both methods can be fairly intensive depending on the level of detail and accuracy required.

Mapping Unexploded WWII Bombs in Germany

While visiting our German business partner, GEOSYSTEMS, in Munich this week we had the opportunity to discuss a very interesting workflow they have been extensively involved with over the last several years.

Dealing with unexploded munitions has remained a challenge for Germany since the end of WWII. Here is a great article outlining the challenge.

However, mapping based on aerial photography has been a success in many regions throughout the country. The workflow involves processing legacy aerial photography taken during or shortly after the war. The imagery is often of poor quality, and may even be lacking the fiducial marks required to establish interior orientation. GEOSYSTEMS tackled this by building an application to reconstruct fidicual locations. After that, they run through the classic photogrammetric workflow and produce stereo images and digital orthos. This enables both 2D and 3D workflows for capturing the location of bomb craters. After munitions locations have been mapped, the data can then be entered in a GIS. Next comes the practical application: in the areas that have implemented this workflow, the database is checked prior to new construction. This helps uncover the potential locations for unexploded munitions prior to construction - which is a life-saving application.

Here is an example of the 2D workflow: vectors of the bomb locations are collected in IMAGINE:


Files with the XY coordinates of the potential bomb locations can also be created:


For the 3D stereo feature extraction process, here's another example depicting stereo extraction in Stereo Analyst for ArcGIS:

KML 3D Buildings in Los Angeles

I thought I would make the buildings I extracted for the 3D City Construction webinar last week available - you can now download them from here.

I'll talk about the tools used to create these another day, but this allows you to take a look at a photogrammetrically-derived quick and dirty city model. It doesn't have all the buildings and they're not all perfect, but it does give you an idea of what's possible. These took me about 8 hours to extract - using anaglyph-mode on my laptop (perhaps less time, I didn't time myself). It would be more efficient to use a proper stereo viewing environment but a laptop works fine for smaller jobs.

Note that if you open the file in Google Earth they'll have the floating effect described here on Thursday... So if you want to follow the workflow described in that post you can grab the terrain from here: both the IMG grid and the RRD pyramid layer.

Defining Your Own Raster Basemap in ERDAS TITAN

Something I forgot to mention in yesterday's post is the ability to define your own raster basemap in TITAN. So what does this mean???

Most virtual worlds have pre-cooked imagery that they serve up as their "skin of the earth" basemap. But what if I want to define my own default imagery, just for my own personal use? Some applications allow loading imagery as layers, but one of the useful features of the TITAN Client is that you can use your own imagery as the default skin.

So how do you do this?

First, load your imagery in TITAN's Geospatial Instant Messenger. The screen capture below shows an ECW orthomosaic that I have loaded.
Next, right click on the layer and choose the "Copy WMS URL" option.

With the URL copied, navigate over to My Services, where you'll see a few difference service options. For the next step, double-click on the WMS service.
This opens up the "Select Services" interface.
Here you can paste in the URL of your image that you copied earlier. Accept it and you'll see the WMS show up in the list of services, displayed below.
Right click on the newly-added WMS service (displayed above) and choose "Set as Default Basemap". You should see it turn red, indicating it is the default basemap.

Finally, fire up the TITAN Viewer. You can see the results of the example I ran through below. Note that my only layers are a terrain layer (see yesterday's post) and a KML file. I don't have any image layers, as my orthophoto is being used as the skin for the virtual world. It's a nifty workflow and can be useful if you operate in the photogrammetric world and create/display/use your own orthos or if you purchase orthos and just want to use them locally as a basemap, without having to deal with layer management.

3D Buildings and Terrain in Virtual Worlds

One the the problems I've grappled with in the past is the challenge posed by modeling accurate photogrammetrically extracted 3D buildings in virtual worlds with a less accurate base terrain layer. The folks over at the Bluesky blog have outlined the same challenge in this post. Here's an example of the problem: when I import my KML 3D buildings into Google Earth, they float off the ground due to the accuracy of the base terrain layer. Here's a screen capture depicting the problem:
Even if you modify the terrain exaggeration the buildings still float. There's a few different methods for resolving this. One method suggested in the Bluesky blog post is to model terrain and imagery right into the KML file and then load it up in GE. This certainly works. Another method would be to take a look at a different system for visualizing the data.

The latest update to ERDAS TITAN has a good technique for sorting out this problem: it allows you to specify your own terrain layer. Once you add your data as layers in the viewer, you'll see a similar problem as depicted above - again due to the accuracy of the default terrain layer. The only difference is that instead of seeing floating buildings the buildings are embedded under the terrain layer, so that some of the smaller buildings are not visible. Here's a screen capture that displays the issue:

You can see that some of the buildings in the foreground appear very "flat", and there are others that are not even visible.

The solution is to add a terrain layer. In this case I processed my own digitial orthomosaic along with a terrain layer during the photogrammetric processing part of the project. I added the digital elevation models as layer, right-clicked on it and chose the "use as terrain" option.
The layer gets shifted to the terrain folder and the TITAN viewer updates to reflect the new base terrain. The result is that the buildings sit perfectly on top of the terrain. Here is the result, from the same perspective as the screen capture above:

In summary, take a look at TITAN if you're interested in this kind of workflow. The ERDAS TITAN Client is free, so you can download it from the ERDAS site and give it a whirl with your data.