Keywords

1 Introduction

Digitization is now a fundamental step in the study and preservation of our cultural heritage (CH). The most common digitization product of tangible cultural assets is a 3D model, while the primary phase is the acquisition of the asset, whether it is a very small one or an entire archaeological site. The need to preserve a CH asset is directly linked to its documentation both for reasons related to its study and for reasons related to its promotion. One of the most important tasks at the stage of analyzing and documenting a monument is its digitization, i.e., the systematic recording and visualization of the elements that define the geometric form and the position in space of the individual parts of the monument at a given moment in time.

In the strictly scientific sense of the term, digitization is a very specific process of the geometric representation of a real object, through observations made by measuring systems, using some suitable computational model which will return, in addition to the representation itself, and the possibility of assessing the accuracy and precision of the deliverable. This representation of the object being captured must end up in a medium of communication (eg a 3D model) to be usable by all the stakeholders.

Digitization, as a process of representing a real object, depends on the accuracy and precision in determining the geometric elements of the object in the single three-dimensional space and the density (or spatial resolution) of the points that will be represented to reconstruct its geometry. The combination of these three parameters (accuracy, precision and density) that determine the digitization, determines the quality of the representation. The type of surveying can be determined by the technique to be applied. The choice of the appropriate technology depends on a number of factors, which must be taken into account in choosing the procedure to follow. These factors concern the characteristics of the object to be surveyed, such as its shape, size, geometric complexity, surrounding space, cost and available time, and the possibility of applying new technology to the process [1]. The traditional techniques that are applied for the surveying of objects or large-scale sites are the topographic and the photogrammetric method.

The development of technology in the imaging sciences in recent years has introduced new techniques for measuring space and objects. The introduction of laser technology at the beginning of the 60s and the understanding of the advantages of the characteristics of the beam it uses, such as monochromaticity, good alignment and the ease of shaping the laser beam, led from a very early stage to its application in measurement technologies and imaging. In recent years, the surveying of cultural heritage sites with 3D laser scanners has been added to classic methods. This chapter presents the modern methods and technologies for the 3D acquisition of cultural heritage assets and addresses the factors observed in the fieldwork that affect the quality of the final product.

2 Technologies for 3D Acquisition

Scanning instruments are distinguished, depending on their nature, into contact and non-contact. Contact scanners measure the object through physical contact, giving relative coordinates to an embedded recording system. These systems are not used usually in cultural heritage objects but mainly in industrial applications. The reason they are not used on cultural heritage objects is their main disadvantage, that is, they require physical contact with the object, which can lead to the change or even the destruction of the object. Heritage items are very sensitive and therefore this method is not suggested.

On the contrary, non-contact instruments and methods are suitable for acquiring cultural heritage assets, as long as they do not emit a beam that can harm the asset. Figure 1 depicts the taxonomy of non-contact instruments used today to acquire cultural heritage objects. Choosing the correct method depends on factors, most importantly the size and complexity of the object. The authors in [1] clarify the term complexity of an object and argue that choosing the correct method is based on the complexity of the object and the complexity of the whole process. Such factors that influence the selection of the correct method are the availability of equipment, the expertise of the personnel and the budget. In the literature, and especially in large-scale digitization, we find the fusion or integration (see Fig. 2) of these methods [2,3,4,5,6].

Fig. 1.
figure 1

Taxonomy of technologies for 3D acquisition of cultural heritage assets.

Full size image

2.1 Photogrammetry Methods

Photogrammetry is the art, science, and technology of extracting 3D information from images. The process involves taking overlapping images of an object, structure, or area and converting them into two- or three-dimensional digital models. The basic principle of photogrammetry is if the directions of the same objects photographed from two extremities of the measured base are known, their position can be located by the intersection of two rays to the same object.

There are two main types of photogrammetry: aerial and terrestrial or close-range. Aerial is the process of using an aircraft to capture aerial images that can be converted into a three-dimensional model or digitally mapped. The most common type of aerial photogrammetry is photography by a suitable camera mounted on an Unmanned Aerial Vehicle (UAV). UAVs have facilitated the safe aerial photography of hard-to-reach objects or hard-to-reach areas where traditional surveying could be dangerous or impractical. Terrestrial or close-range photogrammetry is when images are captured using a hand-held or tripod-mounted camera. The result of this method is usually the construction of 3D models of smaller objects.

The main advantage of photogrammetry methods is that they are cheap, but the disadvantage is that require a lot of processing of the images taken to create the 3D model. However, the evolution of computing power along with the evolution of computer vision algorithms have led to the possibility of an accurate 3D model in just a few hours even for large-scale sites [7]. Along with acquiring images, it is necessary to acquire control points that will help to “stitch” the images and create the model to the correct scale.

Fig. 2.
figure 2

Point cloud delivered by integration of technologies in “DigiArc” project (https://www.digiarc.eu); UAV photogrammetry (top), Terrestrial Laser Scanning and Handheld Scanning (coat of arms).

Full size image

2.2 Laser Scanners

3D laser scanners are active imaging instruments that provide real-time coordinates of the object being captured in three dimensions. Their basic operating principle is that a laser beam is emitted and for each point in space its coordinates (XYZ) are recorded. At the same time, the scanner records the reflectivity of the point giving an intensity value (i). While scanning, modern scanners also collect images with their built-in camera to define the color and texture of the imaged surface.

Fig. 3.
figure 3

Terrestrial Z + FFootnote

https://www.zofre.de/en/laser-scanners/3d-laser-scanner/z-F-imagerr-5016.

laser scanning in “DigiArc” project (https://www.digiarc.eu).

Full size image

A scan can be completed in a few minutes and give us a detailed point cloud that forms and renders the geometry of the object in great detail. All laser scanners work via line of site. This means that on a typical project multiple scans need to be taken from different vantage points to ensure a complete dataset. A change in the position and orientation of multiple scans must be applied so that each point cloud uses a common homogeneous coordinate system of reference. This process is characterized as cloud alignment or registration. This is done by using artificial targets (most common are checkerboards or spheres) in the scans or by allowing enough overlap in the scans to register by recognizing common features (cloud-to-cloud).

3D laser scanners are classified in airborne (known as LiDAR-Light Detection and Ranging), terrestrial (see Fig. 3) and mobile. Airborne scanners are equipped with sensors to measure the position, orientation and altitude of the aircraft during data collection, through inertial sensors [8]. Combining these measurements with distance data collected by the laser scanner produces a 3D point cloud representing the topography of the ground, similar to that generated by a terrestrial laser scanner. Mobile laser scanners are distinguished into those that are vehicle-based and those that are handheld or backpacked [9]. The latters find application in the acquisition of large archaeological sites where, due to the peculiarity of the terrain, the use of a vehicle is impossible.

Based on their operating principle, laser scanner systems are divided into distance measurement scanners and triangulation scanners. Scanners of the first category can measure distance in two ways: a) Time of Flight (Tof) where a laser beam is emitted at the object and the distance between the source and the reflected surface is calculated from the travel time of the signal between emission and reception, and b) Phase Comparison where the emitted beam is converted into a harmonic wave and the distance is calculated from the phase difference between the transmitted and received wave. Triangulation scanners operate under the principle where a transmitter sends the laser beam at a specified variable angle from the edge of the mechanical base to the object and a camera at the other end of the base locates the laser spot on the object. The three-dimensional position of the object is carried out by solving the triangle formed. Triangulation scanners are only applicable to close distances and small objects as the accuracy of the distance between instrument and object depend on the square of the distance.

2.3 Handheld for Close-Range Acquisition

These are portable, small-sized devices that scan small objects with high precision and use the principle of triangulation or the principle of structured light. At the same time, passive receivers capture color data of the object in order to create a more complete model. Scanning with handheld scanners is usually aided by marking on the object certain points with the help of self-adhesive reflective tapes (as with the use of artificial targets).

Structured light is a technology where an infrared emitter emits structured light, i.e. a pre-designed pixel pattern. This is non-visible light (near-infrared wavelength), which passes through a filter and is scattered into a semi-random but constant pattern of small dots projected into the environment in front of the sensor. The depth sensor receives the reflected pattern and calculates the shape and position of the object. The combination of depth information and color information from the camera is combined to create the point cloud.

2.4 Supplementary Instruments

Using control points in a survey can serve two purposes. The first is to provide a network of precise points so that point clouds can be successfully identified in a common coordinate system. Registration using cloud-to-cloud techniques can be very successful, but around a larger structure, for example, or where it is difficult to maintain a good overlap, control points are essential. It also provides certainty and, with some redundancy in the network of points, an estimate of overall accuracy. The second possible purpose of control points is to relate the network of points for the object being surveyed to either a wider coordinate system or a known coordinate system and altitude data.

A very accurate coordinate system can provide the basis for long-term tracking or structural analysis. For large-scale sites, a network of points associated with a known coordinate system can also provide tracking service over a wider area. The georeferencing of the surveyed monuments can lead to studies of their spatial relationship and a more extensive archaeological analysis. An important aspect of the data collection process is defining a network of control points to which all other field metric data can be referenced. The specifications set are to require full georeferencing in a known coordinate network throughout the site. Also, control points are very likely to be needed locally to identify scans, for example, of the interior and exterior of a structure.

To create a network of high-precision control points, measurements are required mainly with two instruments: a Global Navigation Satellite System (GNSS) that works with the principle of Real Time Kinematic (RTK) method and a geodetic station (Total station) (see Fig. 4). RTK is a kinematic determination, in which two receivers (base - rover) are used, provided that there is communication between the receivers, which is carried out either with a UHF modem or a GSM modem. The mobile receiver continuously receives corrections from the base, which has known X,Y,Z coordinates, and uses them to resolve errors from the satellite imagery to achieve greater horizontal and altitude accuracy at the measurement locations. With this method, an accuracy of the centimeter can be achieved almost in real-time. A Total Station is a topographic instrument capable of measuring angles and distances. It combines in a single device, a digital theodolite with electronic distance measurement (EDM) to measure both vertical and horizontal angles and the slope distance from the instrument at a specific point, and an onboard computer for data collection and performing triangulation calculations. It allows its user to collect all the measurements, especially control points in remote areas, necessary for a topographic survey using digital technology.

Fig. 4.
figure 4

Measurements with Total Station (left) and GNSS (right) in “DigiArc” project (https://www.digiarc.eu).

Full size image

3 Factors Affecting the Quality of 3D Acquisition

The quality of surveying and their final product depends on a range of parameters that do not only concern the accuracy of the instrument and the measurements. Factors such as the environment, the characteristics of the object to be scanned, and the methodology followed in each surveying process can affect the quality of the final result of the survey [10]. These factors can be distinguished (see Fig. 5) into those related to the operation of the instrument, the form and nature of the object, the environmental effects and the choice of methodology during the measurement process [11, 12].

Fig. 5.
figure 5

Taxonomy of factors affecting the quality of the 3D acquisition of cultural heritage assets.

Full size image

3.1 Instrumental

These factors depend on the type of instrument and mainly affect the accuracy of the calculation of the measurements of distances and angles of the laser beam. The dimensions of the laser beam spot greatly affect the resolution of the generated point cloud and the determination of the position of the points. The greater its width, the more chances there are to create discrepancies in locating the coordinates of the points to be captured. The shape of the laser beam affects in the same way, which tends to grow at a distance from the instrument [13] and presents a dispersion effect. This factor is known as “beam divergence” and it is a specification provided by the instrument’s manufacturer. When surveying, the specification of beam divergence must be considered when positioning the instrument (see Fig. 6). New instruments tend to allow multiple measurements of a point leading to better measurement.

Fig. 6.
figure 6

Beam Divergence.

Full size image

Another factor that can affect the quality of the point cloud and is directly related to the laser beam is the edge effect. This factor is observed on the edges of the captured objects. When the laser beam reaches a point that is at the edge of the desired object, only a part of it is reflected and returned to the scanner. The remainder may be reflected from a surface behind the edge, if present, from an adjacent point, or not at all. Thus, signals from different areas return to the scanner. Defining the position of the point is done by averaging these measurements, causing the edge of the object to appear in the wrong position. The error, in this case, can range from fractions of a millimeter to a few centimeters. The latest updates of software packages that accompany the instruments can correct some of the edge effect mismeasurements.

Most scanners use mirrors to deflect the laser beam in a specific direction. Any deviations created in the angle created by the emitted beam and the mirror surface can lead to the calculation of incorrect point coordinates. Errors related to the geometric stability of the scanner axes also lead to the definition of the wrong point position. The geometry of the scanner is described by three axes, the vertical, the horizontal and the alignment axis. The vertical axis is the axis of rotation of the scanning head, the horizontal, the axis of rotation of the mirrors and the third is the axis defined by the laser beam. These axes if are not perfectly aligned, errors may occur in the measurements of the angles formed by the alignment axis and the horizontal (alignment error) and by the horizontal and vertical (horizontal axis error). The instrument has a specific time of operation before calibration is needed. Referring to the manufacturer’s specifications can indicate the time that the instrument needs calibration.

3.2 Object Materials and Geometry

The different characteristics of the surface of the objects to be acquired can affect the accuracy of the point distance measurements, e.g. from the scanner, which depends on the reflection of the laser beam. Depending on the material, color, roughness, temperature and humidity of the surfaces of the object being acquired, the reflected beam shows variations. For example, on surfaces characterized by high reflectivities, such as white, distance measurements are more reliable than those with less reflectivity, such as black. This happens because white surfaces absorb less of the visible radiation and thus the intensity of the reflected ray is high, in contrast to black surfaces, which absorb more of the radiation, and the intensity is weak.

If the reflectivity of the object is too high or too low, measurements may not be made or show unacceptable accuracy. Corresponding error results also arise in cases of translucent material surfaces, in which the refraction of the laser beam and its reflection on the surface itself is observed (see Fig. 7). Refraction is the change in direction of propagation of a wave when the wave passes from one medium into another and changes its speed. Reflection of light occurs when the waves encounter a surface or other boundary that does not absorb the energy of the radiation and bounces the waves away from the surface.

Fig. 7.
figure 7

Reflection and Refraction effect.

Full size image

Other object characteristics that affect the results of point distance measurements are size, curvature, and orientation. A typical example is when we have surfaces where the effects of refraction and reflection are observed, leading to incorrect measurements [14]. Optical properties of materials such as glass windows or water (e.g. in a fountain) are a common aspect that affects the quality of the distance measurement, due to the refraction effect. The surface reflection on a mirror of a laser beam normally causes reflected beams in many directions. A solution is to cover these surfaces during the surveying process.

3.3 Environmental Conditions

Environmental factors such as atmospheric temperature, pressure, rain, humidity, and vibrations affect the accuracy of measurements. The presence of wind can cause the presence of dust, which can lead to incorrect measurements whether it is a laser scanner or photogrammetric techniques. For example, temperature changes during the process of a survey affect the speed of the laser beam, which is highly dependent on the density of the air. This results in reducing the quality of its measurement. Also, possible radiation interference from external light sources, such as a projector or sunlight, can alter the measurement results. In the case of aerial photogrammetry, the correct time of day must be chosen so that there is adequate lighting, as shadows greatly affect the photography, and thus the final result. Another factor that affects the measurement results is the movements of the instrument in relation to the object being captured. Throughout the scan, the scanner must be stable to avoid erroneous data collection. Photographing an object with a drone can be affected by the presence of wind. All technologies are accompanied by technical specifications concerning environmental conditions, which must be taken into account and strictly observed during the surveying process.

3.4 Human Factor

Even if the three previous categories are considered, the human factor always comes into the "equation" of quality. The personnel expertise and therefore the methodology they will follow can affect the quality of the deliverables. These factors are related to the methodology chosen either in the phase of defining the specifications during the planning process or the surveying process and in the processing of the generated data. Such factors can arise from a poor choice of settings such as the desired resolution of the sampling and the distance from the object to be acquired or can result from the wrong approach to georeferencing the point cloud. Thus, during the planning of a survey, critical questions must be answered, such as choosing the correct technology or a combination of technologies. For the case of scanning with a scanner, the correct resolution, the use or not and the correct placement of artificial targets, the overlap or not with a percentage of at least 30% must be taken into account for the correct registration of the point clouds with the cloud-to-cloud method and the placement of control points to control the quality of the produced product. In the case of photogrammetry, the correct choice of camera and lens, the desired resolution of the images, the flight plan, the correct orientation of the camera (oblique, nadir), the overlap of the photos at least 80% and the use and the placement of control points to assist further processing and control the quality of the product produced [15].

4 Conclusions

This chapter presents the latest technologies and systems available for the 3D acquisition of tangible cultural heritage assets. Each of these technologies has advantages and disadvantages, while many times in the surveys, especially of large-scale sites, a combination of these technologies is required. Choosing the correct method for a survey depends on factors such as equipment availability, personnel expertise, budget, and item complexity. However, despite the development of technology, many factors can affect the quality of the acquisition and thus of the final product, which must be taken into account during the planning and the process of surveying, so that the metric information is accurate. Along with the evolution of the instruments, the accompanying software for processing the received data is also evolving. These software packages, if updated, can algorithmically reduce many of these factors that affect the quality.