1. Principle of 3D laser scanning
Sometimes we need to analyze a real world object to gather the data of its shape, appearance (color, texture), or dimensional measurement. In order to do that we need a 3D scanner. With 3D Scanner we can gather the data. The data then can be used to reconstruct digital 3D model useful for many different applications. 3D reconstruction used extensively in the production of movie and game, as well as industrial design, quality control, forensic, reverse engineering and
prototyping, computer vision and documentation of cultural or heritage artifacts.
Now the principle of 3D scanning will be explained. The device called 3D scanner will be used to create a point cloud of geometric samples on the surface of the object. These points then will be used to extrapolate the object shape (this is termed 3D reconstruction), the point can also be consisted of the color information, in this case the color on the surface of the object can be determined as well.
3D scanner collects information about surface. 3D scanners keep and collect
the distance information about surfaces within its field of view. The image gathered by 3D scanner describes the distance to a surface at each point in the picture. If a spherical coordinate system is defined in which the scanner is the origin and the vector out from the front of the scanner is φ=0 and θ=0, then each point in the picture is associated with a φ and θ.
Together with distance, which corresponds to the r component, these spherical coordinates fully describe the three dimensional position of each point in the picture, in a local coordinate system relative to the scanner. In many cases, single scan from one angle will not generate a complete model of the object. We need to take multiple scans on multiple scans from different angles to gather information about all sides of the object. These scans then transformed into some reference
system and a process termed image registration will be applied to obtain a complete model [wikipedia].
A scanner will emit laser and detect its reflection in order to probe an object. There are few types of laser scanner, two famous laser scanner type are time of flight laser scanner and triangulation laser scanner.
The first type of 3D laser scanner is the time-of-flight 3D laser scanner. This type of active 3D scanner utilizing laser light to probe the object. The principle behind this laser scanner type (also known as laser range finder) is using laser to emit a pulse of light and calculating the distance between the scanner and the surface of the object by measuring the time of the round-trip-time of a pulse of light. Since the speed of light c is known, the roundtrip time determines the travel distance of the light, which is twice the distance between the scanner and the surface. If t is the round-trip time, then distance is equal to ct/2.
The accuracy of a time-of-flight 3D laser scanner depends on how precisely we can measure the time: 3.3 picosecond approximately is the time taken for light to travel 1 millimeter. The laser range finder only detects the distance of one point in one direction (the direction of its current view). Therefore, the scanner scans its entire field of view one point at a time by changing the direction of range finder’s to scan different points. The direction can be changed by either rotating the range finder itself or using rotated mirror systems. The rotated mirror methods are
faster and more accurate because mirrors are lighter. Generally time-of-flight 3D laser scanners can measure the distance of 10.000-100.000 points per second.
This type of laser scanner is more suitable to scan a far object, like buildings, rock formations etc produce a 3D model. [Wikipedia]. One of application of this scanner is Lidar (light detection and ranging) scanner, the example of reconstructed object using this type of laser scanner is shown below :
The second type of laser scanner is triangulation laser scanner, it is also a active scanner that uses laser light to probe the environment. These laser scanners shines a laser on the object and take advantage of a camera to look for the location of the laser dots emitted. Depending on how far away the laser strikes a surface, the laser dot appears at different places in the camera’s field of view as shown below :
This technique is called triangulation because the laser dot, the camera and the laser emitter form a triangle. We get three informations :
1. The length of one side of the triangle, the distance between the camera and the laser emitter.
2. The angle of the laser emitter corner.
3. The angle of the camera corner (determined by looking at the location of the laser dot in the camera’s field of view).
With three information described above, we can entirely establish the shape and size of the triangle and gives the location of the laser dot corner of the triangle.
The triangulation process can be described as follows. A laser stripe (or line) approximately 2 ½ inches long is projected from the laser head onto the surface to be digitized.
The CCD cameras,at a known distance from the laser, then capture the light from the laser beam as it reflects on the part. Through trigonometry, the XYZ co-ordinates of the stripe can then be calculated.
Depending on software settings and the sensor used, up to 650 individual data points can be gathered from a single laser line. The object that was scanned will now be represented by these points, commonly referred to as a 'point cloud', numbering from a few thousand points up to possibly a million or more.
The number of points scanned on a part will depend on the size and amount of detail the part has. The more detail, the higher the number of points necessary to describe the part. [Steinbichler].
In recent development of this type of laser scanner a laser stripe, instead of a single laser dot, is scrutinized across the object to speed up the acquisition process.
Both types of scanner have its own advantage and drawbacks. They used on different purpose and situations. The advantage if time-of-flight laser scanner is that it is has a capability to scan a very long distance, sometimes in order of kilometers. This scanner is suitable for scanning large structure like building or geographic characteristic, the drawback of time of flight laser scanner is their accuracy. This is because of the high speed of light, making the round-trip time is difficult to calculate in the order of millimeters. With time of flight scanners accuracy is
decreased when the laser hits the edge of an object because the information that is sent back to the scanner is from two different locations for one laser pulse. The coordinate relative to the scanners position for a point that has hit the edge of an object will be calculated based on an average and therefore will put the point in the wrong place.
On the other hand, the advantage of triangulation laser scanner is its accuracy. They can only scan in limited range (order of meters) but their accuracy is in the order of tens of micrometers.
That is to conclude that triangulation laser is more suitable to reconstruct a small 3D object. When using a high resolution scan on an object the chances of the beam hitting an edge are increased and the resulting data will show noise just behind the edges of the object.
Scanners with a smaller beam width will help to solve this problem but will be limited by range as the beam width will increase over distance. Software can also help by determining that the first object to be hit by the laser beam should cancel out the second. Laser scanning shortens the digitizing process by collecting the data at a much faster pace than conventional measurement techniques. Scanning with the laser eliminates the issues relating to cosine error, deflection, mechanical probe offsets, and probe size and/or shape.
2. Optimization of 3D laser scanning for reflective surfaceWhen the data from 3D laser scanner are reconstructed, the large data sets usually have to be processed. Hence, it is often necessary to minimize the number of points while minimizing the loss of information simultaneously. Moreover, the generated point cloud usually contains a considerable number of errors. The errors come from the measurement system and the scanned object’s surface. Outliers and other incorrect points are an important factor when discussing precision measurement.
Therefore, they have to be first detected and removed or corrected from the point cloud in order to get a clean model. Hence the data measuring can be done more
precisely.
Based on this fact, optimization of the laser scanning system can be realized if the data quality of each individual point in the data set is improved . This can be accomplished by considering the specific settings of the scanning system, particularly, the parameters of the laser and the camera. Optimizing their positions based on the results of the point cloud analysis results in
better point cloud quality in a successive scan.
There are many sources for incorrect scan data. Based on many numerous example scans, most of the outlier and other erroneous points are caused by reflections. In these cases, the high energy laser beam is reflected from mirroring surfaces such as metallic or glass. Consequently, there are too much light hits the sensor of the camera and blooming effects occur. In other cases, a direct reflection may miss the camera. In addition, a part of the object may lie in the path from the projected laser line to the camera causing a shadowing effect. All these effects are responsible for gaps and holes. At sharp edges of some objects, partial reflections appear. In addition, rough surfaces cause multiple reflections and, therefore, indefinite point correlations.
Other problems are caused by possible range differences originating from systematic range errors resulting from different reflectivity of the surfaces elements. Since the scanner systems are typically used in industrial environments, some atmospheric effects like dust and dirt possibly affect the quality of the image obtained by the camera. Furthermore, aliasing effects in the 2D image processing lead to high frequent noise in the generated 3D data.
Therefore, the resulting 3D point data is noisy and partially erroneous. However, a lot of these errors can be minimized by an optimal alignment of the projection system and the object surface so that as few as possible reflections can appear. In order to arrange the setup properly, the quality of the generated point cloud has to be analyzed and evaluated.
The steps to get optimized 3D reconstruction by improving point cloud quality are :
1. Optimize the projection and viewing conditions. This way, the quality of the point cloud has to be quantified with respect to the position of laser and camera. Improving the recording conditions will result in less erroneous 3D points.
The proposed techniques for improving the recording condition are based on the system’s parameters in 2D (e. g., contrast and line thickness) and in 3D (e. g., camera and laser positions, focus area, etc.) to estimate the quality of each single point [Teusch 2004].
2. After the point clouds is quantified, now we have to correcting a single measurements of a point. The goal is to minimize the number of points and clean the point clouds from noise. The algorithm to do it is by analyzing the points using B-Spline approximations. The point clouds are approximated by a sorted set of B-spline curves for iterative smoothing and closing gaps [Teusch 2004]. The edge information is derived from these curves and then reconstructed by using NURBS curves with respect to quality and curvature of each single point on it. NURBS or
Non-Uniform Rational B-Splines, is a mathematical representations of 3-D geometry that can accurately describe any shape from a simple 2D line [Fisher 2002].
3. Now the triangulation algorithm can be conducted as usual.
The figure below show the example result of this method:
