- SPECIAL REPORTS
- THE MAGAZINE
Tripod-mounted 3D laser imaging systems (or laser scanners) are new, incredible tools. With the unique combination of measurement rate, range and portability, laser scanners are capable of revolutionizing some areas of the surveying industry. Not only can traditional projects be completed more efficiently, the technology has provided a solution for applications that were previously difficult or even impossible using conventional sensors. Based on this information one might assume that data can be collected and processed automatically by the push of a button. While the advantages of the technology are apparent, conducting a survey is not necessarily as straightforward and automatic as it first appears. As with anything else, a successful project will depend on reasonable expectations, adequate planning and an understanding of the principles of the technology. Unreasonably high expectations of the product or underestimating the scope of the project can lead to disappointment on many fronts. When conducting a laser scanner survey, following a certain process will ensure that a complete, accurate and usable end product can be provided in a timely and cost-effective manner. This process requires end users to manage their expectations, understand the objectives, adequately plan the collection of data and establish a thorough plan for data processing. With proper planning and a good understanding of how the technology should be applied, laser scanners can produce results that, depending on the application, will match or outperform their conventional counterparts.
Managing ExpectationsBoth the physical and technological characteristics of a product determine its ability to perform in a given application. As useful as it may be on a construction project, a screwdriver can’t be used where a chisel would be. By the same token, with laser scanners, the end user must manage his or her expectations when preparing for a scanning project by understanding how to properly apply the technology, which includes understanding how to size up the project to determine whether the technology can be used. There is much informal knowledge about laser scanners that can be gained, though. Through advertising, positive feedback and industry buzz, there is an understanding that the technology can collect very fast, accurate measurements over long ranges. But considering whether to use a laser scanner on such information alone is not advised. The potential for disappointment and inadequate results is high when the preparation is low.
For example, a request for an accurate high-resolution model of a dam is valid. However, based on that requirement one must ask the questions: What is meant by accurate? What is meant by high resolution? What do long ranges mean? In each instance the requirement can be quite subjective. The user must ask these questions and educate him or herself as to what is and what is not feasible. If a user expects a model with a 1-mm resolution and receives one with 2-mm resolution, he or she will be disappointed. If a 1-mm resolution solid model is provided but is an unmanageable file size, then he or she will also be disappointed. Ultimately, it must be decided, based on the capabilities of the technology, experience of the service provider and requirements of the project what exactly the end deliverable will be. A cost-effective solution must be determined while at the same time fulfilling the necessary obligations of the task at hand without compromising the integrity of the project. Education on the limitations and capabilities of the technology is usually the simplest way to avoid this problem. For example, while a 1-mm solid model is possible, it may not be in the user’s best interest. Based on the application (and the provider’s experience) a 5-mm model may be adequate. It is wise, as when planning all surveying and mapping tasks regardless of the technology being used, to establish a middle ground that meets the user’s requirement yet provides the most cost-effective solution.
Understanding the ObjectiveWhen a proposed solution is accepted, one must then assess the steps necessary to complete the project. This can be approached in three general phases by answering these questions: What is being scanned? What is the final product to be? And what steps must be taken to generate the final product (data collection and data processing)? The last phase can be thought of as preparing a project outline. One must consider all of the key points before moving on to the fine details. For example, a very large area may require a scanner with significant ranging capability while a small, enclosed area may require a large field of view. Accuracy and resolution requirements will help determine the length and possibly the range of the scans. Armed with this information, it is possible to estimate about how long the scanning will take and how much data will be collected. The amount of data, the number of scans and setups, and the format of the deliverable will dictate the direction of data processing. Different software packages may possess capabilities that are critical to the completion of the project. One should use the project details to determine what software packages will be necessary and make a broad estimate about how many processing hours will be involved.
Project PlanningBefore taking scanning equipment into the field, it is beneficial to first “scout” the area. A preliminary reconnaissance will identify potential problems that may cause a delay when scanning starts. This is especially important if the project is in a remote area. It is extremely frustrating to arrive on site and discover that targets (for alignment or geo-referencing) or external power will be required due to unforeseen circumstances. When scouting the project for the first time, one should keep the required accuracy and resolution of the project in mind. Based on the size and complexity of the area this will help determine the length and number of scans that will be necessary to complete the job. A number of potential problem areas of laser scanning are listed below. Depending on the characteristics of the survey site, any or all of these may be issues when data collection commences.
Accessibility refers to either the level of difficulty to obtain permission to enter a project area or the potential difficulty in reaching a required scan location. In most cases, permission to enter the project site will not be an issue. However, access to potential scanner locations may be logistically or physically difficult. For example, the survey of an airport tarmac may call for the elevation of the scanner to facilitate data collection or call for closure of a portion of the area. In areas where access to a building roof is available this is not an issue. However, in some cases external platforms or elevation equipment may be necessary. If this has not been planned, delays will occur. Also, many projects may require the user to hand carry the equipment over long distances or difficult terrain. If the system (plus accessories) is too large, this can be an obstacle. In this case, a plan must be made to have additional personnel on hand or, if possible, scan from a different location.
Most laser scanners offer the ability to operate from battery or AC power. In remote locations one must be sure to either bring enough batteries for a full day of scanning or prepare an alternative power source. Some vendors offer solar power or the option to operate from a vehicle battery. If it is common to scan in isolated or out-of-the-way areas a portable generator may be a viable option. Preparing an adequate power supply will ensure that no time is lost due to a lack of electrical energy to operate the system.
Aligning Multiple Scans and Geo-referencing
There are many software packages that offer the ability to align multiple scans together without the need for targets. However, to work effectively, most require a minimum of 10-15 percent overlap between adjoining scans. Depending on the circumstance this may not be possible. In scanning a bridge, for example, it may be difficult to collect enough overlap data to align scans on the top of the bridge with those on the side without the use of targets. If not identified very early a second trip may be required to collect more data when scans do not accurately align during processing. Similarly, targets may be required if it is necessary to transform the point cloud into any number of grid systems. Generally speaking with a scene that is geometrically complex, targets may not be required. Identifying and locating features in the point cloud may be done simply with a reflectorless total station or an RTK GPS pole. This is typical in architectural or as-built environments. However, with many projects distinct points are not always available. For example, when scanning a rock face in an open pit environment there is generally nothing distinct to base a grid transformation on. In these instances targets may have to be placed in the scene to allow for accurate coordinate transformation. In situations like these it will be necessary to strategically place targets in areas that are identifiable in all scans that are to be aligned and do not have adequate overlap or require geo-referencing.
While “time-of-flight” laser sensors collect accurate ranges they do not possess the ability to see through physical surfaces (there are a small number of exceptions but by and large this is the case). Obstructive objects will produce a “laser shadow” in the data. The shadow is an area where no data was collected due to the obstruction. To account for this, one must scan from multiple perspectives to eliminate as much shadowing as possible. One must be extremely aware of this while in an as-built environment. In some instances a laser shadow could potentially “hide” an entire pipe or beam. When scanning an unobstructed flat surface, laser shadowing is a non-issue since there is nothing impeding the laser measurement. Undulating or complex surfaces will require multiple scans to minimize laser shadowing. Depending on the requirement, interpolation in the shadowed area is possible. However, if full detail is required on a complex surface, one must plan for more scans, more scanning time and by definition more data.
If the object that is being scanned is 500 m away, it is not feasible to use a scanner that has a maximum range of 100 m. On that basis, one must calculate ranging capability that will be necessary during the project. Setting up farther away from the subject(s) will allow for a larger area to be covered faster (due to a larger field of view) but will reduce the resolution and, to a certain extent, the accuracy. The closer the subject, the more scans will be required to cover the same area. If targets are necessary for each scan the setup procedure will be very time-consuming. It is also possible that practical limitations may force the collection of longer-range data. Surveying a bridge that crosses a large body of water may force a long-range scan to be collected from the shore. When scouting the scan site, identify the practical limitations of the available hardware and consider obtaining substitute hardware if the circumstances warrant it.
In every case public and worker safety is of paramount importance during a project. When working with lasers of any kind the subject of eye safety is inevitable. Since it is nearly impossible to ensure that no one steps into the field of view of the laser (especially at long ranges), the provider must be sure to use an eye safe laser. Most available scanners are eye safe, although some are only eye safe under specific circumstances. When data is being collected in an area where people will routinely be in front of the laser while scanning, a Class 1 infrared laser should be used. Class 1 lasers are safe under reasonable, foreseeable conditions of operations, including the use of optical instruments for intrabeam viewing.
Post-processingData processing will account for the largest portion of the labor required to complete a project. Depending on the end product, data processing will represent 50-95 percent of the gross labor. As such, establishing a proper data flow and correctly budgeting labor hours will be vital to the success of any scanning project. Exceptional data may be collected but if processing does not produce a viable product, it becomes worthless. Laser scanners can collect many gigabytes of data. In its rawest form, very few end users would be able to efficiently extract the necessary information. Due to this limitation most users will require either a reduced point cloud, CAD drawing or solid model.
When preparing for data processing there are several important questions that should be addressed. The format of the final product is the most important consideration since it will directly dictate the amount of processing that is involved as well as the software that will be necessary. There has been a misconception that a solid model must be generated to obtain any useful data from a scan. In fact this is not the case. In some instances a simple point cloud may be sufficient. In other instances CAD drawings may be in order. Ultimately, this needs to be established early in the scanning process, keeping in mind the time and skill necessary for processing and exporting into different formats. For example, delivering a finished CAD drawing or solid model will naturally require more processing time than providing a simple point cloud since more processing steps and manual edits are involved. The creation of a CAD drawing requires the processing of a point cloud followed by the extraction of features and lines to a CAD format. When collecting data one must keep note of the file sizes that are being recorded. Extremely dense data can be useful but it will also require very powerful computing hardware and will increase the processing time significantly. While it is recommended that the user ensure sufficient coverage for the job with the possibility of collecting extra collateral data for future use, one should keep file size in mind and keep laser shot density to a manageable level.
Unleashing New MethodsRegardless of the tools, no project can be successful without proper planning and a clear objective. Laser scanning projects are no different. Managing the workflow not only increases efficiency, but it can also identify potential problems and limitations before work is well underway. Early identification of what is required and what form the solution should be in will also lend itself to competent and well-organized data management and manipulation. By establishing objectives and understanding the data collection and data processing that will be involved, one can make an accurate assessment about what is realistically achievable and the time it will take to complete it. The cost of the project will depend on this.
Laser scanners are wonderful, revolutionary tools, but they must be put into a position to succeed. Following a tight procedure will ensure a unique solution that is, in many cases, far more practical then anything currently available.