Differentiation of 3 D scanners and their positioning method when applied to pipeline integrity

In the world of 3D scanning, the right scanner depends on the application and the main goal of the people who will use it. Each method has its benefits but also its trade-offs. This paper aims at helping service providers and asset owners select the most suitable 3D scanner solution for their inspection needs. 3D scanners are mainly differentiated according to their positioning method. The measuring arm, tracked 3D scanner, structured light, and the portable 3D scanner categories will be investigated. More specifically, the two main positioning methods used by portable 3D scanners will be discussed: positioning through targets, and positioning through natural features. A third method called hybrid consists in combining the two. 3D scanners are used for pipeline fitness-for-service evaluation in replacement of conventional methods such as pit gauges and ultrasound probes. Corrosion and mechanical damage can now be characterized with very high accuracy and repeatability. Each scanner category has been tested for corrosion assessment on a pipeline. We will see how they perform against each other and the importance of a proper positioning method. 11th European Conference on Non-Destructive Testing (ECNDT 2014), October 6-10, 2014, Prague, Czech Republic M or e In fo a t O pe n A cc es s D at ab as e w w w .n dt .n et /? id = 16 31 7 1. Bridging physical and digital worlds 3D scanners are tri-dimensional measurement devices used to capture real-world objects or environments so that they can be remodeled or analyzed in the digital world. The latest generation of 3D scanners do not require contact with the physical object being captured. 3D scanners can be used to get complete or partial 3D measurements of any physical object. The majority of these devices generate points or measures of extremely high density when compared to traditional “point-by-point” measurement devices. 1.1 How 3D scanning works There are two major categories of scanners based on the way they capture data: • White-light and structured-light systems that take single snapshots/scans • Scan arms and portable handheld scanners that capture multiple images continuously. Scanning results are represented using free-form, unstructured three-dimensional data, usually in the form of a point cloud or a triangle mesh. Certain types of scanners also acquire color information for applications where this is important. Figure 1: Triangulated mesh representation Images or scans are brought into a common reference system, where data is merged into a complete model. This process called alignment or registration can be performed during the scan itself, called dynamic referencing, or as a post-processing step. 1.2 3D scanning categories and positioning methods The benefits and limitations of a 3D scanner are typically derived from its positioning method. That's why it is valuable to take a look at positioning methods within the different 3D scanner categories. 1.2.1 Measuring arms, portable CMM scanners CMMs (coordinate measuring machines) and measuring arms can be equipped with either fixed-probe or touch-trigger probe heads. It is also possible to mount a 3D scanning head on a CMM. CMMs with portable arms are positioned using the mechanical encoders integrated in the arm. Many different tools can be mounted on portable CMMs, making it possible to easily integrate scanning and probing in the same project. Portable CMMs need to be fixed on a surface and use a physical link (arm) as their positioning method. This makes them prone to vibrations and other environmental constraints that can affect the performance and quality of the result. They also lack flexibility in terms of the locations in which they can be used and the shape of the objects they can scan. Figure 2: Articulated Arm 1.2.2 Tracked 3D scanners Optical tracking devices can track various types of measurement tools, including the positioning of a 3D scanner. These scanners use an external optical tracking device to establish positioning. They usually use markers such as passive or active targets that optically bind the tracking device to the scanner. Tracked 3D scanners provide very good accuracy and excellent precision throughout the measurement volume. The optical link is strength of this technology and also one of its limitations. The tracker must always have a clear and direct line of sight to the 3D scanner. Trackers often have a limited working volume. Extending the scanning parameters adds complexity to the process and can induce some additional uncertainty in the measurements. Finally, tracked 3D scanners are usually more expensive than solutions such as portable 3D scanners. Figure 3: Tracked scanner, Creaform Metrascan 1.2.3 Structured-light 3D scanners These scanners project a pattern of light onto a part and process how the pattern is distorted when light hits the object. Either an LCD projector or a scanned or diffracted laser beam projects the light pattern. One, two, or sometimes more sensors record the projected pattern. The positioning method between two pictures taken to perform registration is usually done off-line using targets or natural features. If only one camera is used, the position of the projector in relation to the camera must be determined in advance; if two cameras are used, the stereoscopic pair must be calibrated in advance. High-end structured light scanners generate very high-quality data. They typically deliver excellent resolution, which allows for the smallest features on an object to be captured in the results. While white-light scanners can acquire large quantities of data in one scan, overall project speed is not always improved by this methodology. Multiple scans are required in most cases to cover all angles on more complex parts, which is very time-consuming. Figure 4: White-light scanner pattern projection 1.2.4 Portable 3D scanners Many types of portable 3D scanners are available on the market, principally using laser-line or white-light technologies. Laser scanners project one or many laser lines on an object while white-light devices project a light and shade pattern. Both will analyze the resulting deformed projections to extract the 3D data. Handheld scanners rely on two cameras to create what is called stereoscopic vision. This enables the device to determine the scanner position in relation to specific points, which could be positioning targets, the object’s natural features or textures. Some newer portable scanners use a mix of positioning types called hybrid positioning. Portable 3D scanners can be transported with minimum effort and are often easier to use than other scanner types. They can combine multiple positioning methods, providing the accuracy of positioning targets with the flexibility of object features and texture positioning. The most advanced technologies can acquire more than half-a-million points per second and rebuild the 3D triangle mesh live during the scanning process. Handheld scanners do not require a mechanical link or a direct line of sight with a tracker. This enables them to reach narrow and enclosed areas. Portable scanners use self-positioning on a more local area, which means that errors can stack up as the scanning volume grows. It is possible to circumvent this by using technologies such as photogrammetry and positioning targets to minimize errors, but these additional steps might increase setup time and limit the size of the objects or areas that can be scanned efficiently. Figure 5: Portable 3D scanner, Handyscan 700 1.3 Big performance in small packages The knock on portable 3D scanning systems used to be that they couldn't match the performance of their hulking big brothers. That might have been true even five years ago, but it isn't any longer. Consider these specifications for the newest generation of hand-held scanners: • Accuracy up to 0.030 mm (0.0011in) • Resolution up to 0.050 mm (0.0020in) • Scanning speed of up to 550,000 measures per second • Weight of less than three pounds • True portability in the field or the shop • Stability in environments where vibration is the norm. This combination of speed, accuracy, portability and ease-of-use has been delivered through a series of technological breakthroughs, the biggest of which is self-positioning. Scanners can now rely on the object being scanned as the reference for positioning. This is as opposed to using an external positioning device such as an arm, CMM or tracker. True portability means being able to comfortably transport a 3D scanning system to remote locations and operate in tight spaces and unstable environments. But it also means having the flexibility to accomplish multiple types of tasks within a dedicated room or for outside field work. Beyond the ability to take a 3D scanner anywhere, portability has the advantage of flexibility: optimizing a 3D scanning investment, as it enables users to accomplish a wide range of scanning tasks with a single, transportable system. The greater variety of objects you can scan and the greater number of places in which you can scan, the greater the return on investment. 2. Positioning methods for portable 3D scanners Different positioning methods used by portables scanners can have an important impact on performance and usability. In the sections that follow, we'll explore the two main positioning methods for portable 3D scanners positioning through targets and positioning through natural features (geometry), as well as a third method that is a blending of the two. At the end, readers should have a good idea about what type of positioning system best aligns with your 3D scanning needs. 2.1 Positioning through targets In this method, positioning targets are applied before scanning, either on the object or around its immediate surroundings in the case of a very small pipe, for example. The targets enable users to register all the different camera frames for the 3D data sets acquired by the scanner. Targets usually hav