One of the most common misconceptions with laser scanning is that you only need one or two scans to capture an entire room. While this might be true in some cases, it certainly isn’t for the most part. In fact I personally love it when I can get away with one or two scans in a room, it makes me feel like I am getting my job done really fast! I wanted to take a moment to try and show people that may be looking to use scanning how we work out how many scans it takes to do a job so it doesn’t catch you off guard when you get a quote back for 20 scans when you thought it would take 3 or 4.
When most people contact me for the first time for quotes on scanning they commonly say something along the lines of “I only need to scan an area of X x Y, how much will that cost?” Unfortunately it isn’t that easy. To work out how many scans a job will take relies far more on what is in the areas as opposed to how big the area is. we also have to consider things like how much detail you want to see of that area and what you want to do with the scans.
I could write a few thousand words on this topic, but to save everyone the boredom, I will let some pictures do the talking.
Example 1 – The empty room.
In this case, it is easy to see that one scan will do the job, no worries.
But what happens if there is an obstruction in the middle of the room? That should still only take 1 scan right?
Example 2 – 1 Object in the room
Assuming that you want the whole room and the object in it you will only need two scans as pictured below.
This is true if the object is centered and reasonably uniform. There are some other things to consider here though.
When you do more than one scan on a job you have to run the scans through a registration process to align them with each other in order to form one large point cloud. There are quite a few programs around that will do this and they all have 2 basic ways of alignment, using targets or a process commonly called “Cloud to Cloud”. With the former it is a case of placing targets out on site and ensuring they are adequately scanned while on site for the software to pick them out in the point cloud when registering the scans, or using the scanner while on site to acquire the targets. In the case of the latter the software uses large groups of points in the scans to try and align the scans based on geometry. The software companies usually use things like walls, floors and ceilings as examples of what the software tries to use to do the alignments.
There is a hotly raging debate about which is the best process. While this could be an entire article in itself, it really is just a case of personal opinion and personally preferred workflow. The one thing that most people agree on is that the more complex the geometry scanned the worse the results of cloud to cloud. The one thing to know as someone who is looking to get a quote on scanning is that the two processes take different amounts of time while on site and off site.
When using targets it typically takes a little longer on site to set the scanner in a position to get the targets and then a little bit of time to acquire the targets if the scanner can do that. This process on site isn’t required for cloud to cloud, however it can often mean doing more scans on site to get sufficient overlap and there is typically more time spent on ensuring good registrations back in the office. Depending on how your service provider prices their work, this can change the cost of your job significantly. Just keep in mind cheaper or faster is not necessarily an indication of better. This is one of those cases where you need to rely on the service provider to work out what is best for your project.
For what it is worth, I am firmly in the belief that the best way to get reliable registrations is to acquire the targets while on site. As most of our work is in and around processing facilities that have quite complex geometry, I am reluctant to rely on cloud to cloud and use targets for almost all of our work. I have done jobs where I just put the targets out and made sure they had enough points on them to do the target acquisition in the software in the office, to keep the time on site down. But I found when doing this that the scanner does a better job of picking the targets than the software does. While the registrations have always come out well, I prefer to err on the side of caution and do what I feel is the better process as much as possible.
But I digress a little.
In the case of the above example, a cloud to cloud registration would work well due to the amount of overlap in the 2 scans and the simple geometry, but what if there were a small room in the corner of this space?
Example 3 – Multiple Rooms
As you can see below, two empty rooms actually requires 3 scans. This is for both the cloud to cloud and targeted registration workflows due to the small door on the room.
(It is probably worth pointing out here that the scanner is not drawn to scale obviously.)
(In the above example, it would be necessary to do a scan with the door shut then a detail scan over the area of the door with it open in order to get the result shown.)
The only way you could get around this is if it were possible to acquire 2 or more targets at reasonable spacing through the door way, which with personnel doors typically isn’t possible. The general rule of thumb I work on when scanning multiple rooms, like a house for example, is one scan per room and one in each doorway.
So what happens when we want to scan something with a lot of objects in it like a plant room?
Example 4 – The plant room
As you can see in the example below, there are now 4 scans required to scan the same sized area that we have been working in the whole time.
In the image below, if you look closely, there are still two small areas where the scanner hasn’t quite picked up on the back of the round objects on either side of the room.
While this example would be a suitable capture for most jobs, if there was important detail in the areas that were missed it would require at least one and quite possibly two more scans being added just to fill that small area.
So now that you have read all this you know exactly how to quote a scan job and you can start telling your scan service provider how many scans a job will take right?…Wrong!
While these examples ring true for the most part and are a good rule of thumb to working out the amount of scans required, there is still a lot more to working out how many scans it will take to do a job. For example the materials that are being scanned can influence how many scans it will take to complete a job. In Example 2, if the object in the room were made of stainless steel for example, it may be better to do 4 scans instead of 2 to get the best results. In all of the examples shown above it is showing a huge area with smooth walls and no structural steel. While this scenario is possible, if this were a portal frame building (which is far more common) and you needed to scan to capture the steel work, even if the room were empty, it may require 6 or more scans to capture the detail required to remodel the steel work. Add to this the 3rd dimension that we are actually working in and not just the 2D shown in these examples, things like pipe work and other objects blocking lines of sight, working out how many scans a job will take can be a bit of a dark art and one that takes a lot of practice to get good at.
I once had a job that should have only taken 6 scans end up needing 8 because a truck was parked in the middle of the area we were scanning and it couldn’t be moved. So it is always tricky.
The Take Away
So what should you take from this as someone that is looking to get a quote on scanning? While you can use the above examples as a guide to get a slightly better idea of what a scan job will take, it is not gospel. If your scan service provider says it will take more scans than you have worked out, there is probably a good reason for it and you should trust them. Instead of trying to work it out for yourself, the best bet is to give your service provider as much information as you can about the site and what you want to do with the scans as well as the critical details that need to be captured, then let them work it out.
So what information do you need to give your service provider to get an accurate quote? Nothing beats a site walk through if it is possible. But a good set of 2D plans and some photos is second best. Even a rough hand sketch of the site layout and a good description is better than someone with no scanning experience trying to work out how many scans a job will take. Don’t forget to let your service provider know what you intend to do with the scans and if there are any areas of particular interest that need to be captured, like steel work tie in points for example.
If you have any questions or comments I am always happy to answer them. Feel free to ask below or get in touch. Also follow our Complete 3D Concepts LinkedIn page to keep up to date with what’s happening.