Abstract
The constant development and improvement of laser scanning instruments together with the digital rebirth of photogrammetry (photomodelling) has triggered the operative and methodological evolution that in many aspects has already facilitated the acquisition phase, especially referring to the time required for the procedure. The present paper, however, investigates the methodological changes occurred especially in the last decade concerning surveying and its representation. We shall describe the peculiarities and articulations of these methods aiming at pointing out some conceptual issues involving measurement, interpretation and representation. The final objective is to construct an “operative critical method” which seems indispensable to standardize and regulate the procedures of data collection, elaboration and representation of architectural artifact providing results more objective and reliable: in other words more scientific.
TopIntroduction
The possibility of carrying out an accurate survey of complex geometries, a task that would yield a deep study of the object with the perception and analysis of geometric and material details performed at considerable cost with traditional techniques, has recently become a necessity felt stronger with the passage of time. Similarly, it is even more necessary to archive and document the actual state of the objects analyzed, especially when there is a need for reconstruction or maintenance work to carry out.
Precisely because of the typological and dimensional variety of architectural artifacts, their surveying, analysis and representation allows one to ponder equally diverse and variable problems. One of the possible solutions to the problems inherent in the surveying process is working out guidelines that would standardize all the stages of surveying independently from the object under analysis.
In principle, the aim of data acquisition, elaboration and representation is to produce “scientific results”, that is to say, outcomes that are repeatable and similarly more objective. For this reason, the guidelines should focus on the following problems:
- •
Understanding and solving of the criticality level of the acquisition instruments and technologies;
- •
Suggestions for the correct application of the technologies under consideration;
- •
Possibilities of speeding up surveying and representation;
- •
Reciprocal feedback between operative protocols, instruments and the general Survey Theory.
The present article puts forward a “critical operative method” which seeks to establish some common principles of the entire process while not aspiring to provide a formal general protocol. Obviously, the point is to put forward some guidelines which will be dynamic and continuously evolving, conceived with the aim of preserving the characteristics of “scientificity” of the entire process of surveying and representation while at the same time constituting the fruit of constant experimentation with diverse objects with the use of different technology (long range laser scanner, short range laser scanner, photomodelling).
TopBackground
The surveying process has undergone considerable changes with the passage of time. This was due to the introduction of innovative technologies for data acquisition as well as to the application of digital technologies for representation. The new technologies solved some of the problems of surveying, for example, accelerating data collecting, yielding graphic models less subject to interpretation and more adequately corresponding to reality, or offering the possibility to reproduce the graphic models many times with the view to sharing it with other researchers and scholars (see Figure 1, see Figure 2).
Figure 1. Temple of Divine Claudius, Rome, photography.
Figure 2. Temple of Divine Claudius, Rome. Data acquisition and elaborations. On the top: Registration of the 3D scansion. On the bottom: Selections of the reference planes to create the sections.
The technology developed in the last twenty years makes it possible to obtain millions of points almost automatically in ever-shorter spans of time. This performance level produces quite detailed graphic models but they do not go well with subjective interpretation that is the characteristic feature of the stage of representation. What is more, the collected data and the 2D and 3D models – if they are digital - become elements to be “archived” so that they can be used and shared by various users for different purposes (see Figure 3).
Figure 3. Temple of Divine Claudius, Rome. On the top: Proportioning of the complex using the module. On the bottom: Example of an image with the architectural drawings applied as a texture to the 3D model.
Key Terms in this Chapter
Mathematic Model: Model where the shape is described continuously through parametric equation of the surface of which is composes.
Geometric Model: Model without any chromatic characterization and information about the state of preservation of the surveyed object. It is very important and useful to understand the geometry, proportions, measurement and the relative position of the elements.
Stereoscopia: Set of techniques that allow one to record three-dimensional visual information or creating the illusion of depth in an image. The stereoscopic vision is obtained starting from a pair of images taken from two points (separated by a known distance) and presented in a separate way to the vision of the two eyes, thus creating an effect of depth due to the superimposition of the two views.
Computer Graphics: Discipline that studies the techniques and algorithms for displaying numerical information produced by a computer.
3d Survey: Process that involves the passage from the real object to its representation, which leads to the definition of graphic models through a cultural process aimed at a critical reading of the object and its description 3D and 2D models, able to describe the object to the appropriate scale of representation.
Rendering: In computer graphics, the process of producing 2D images from a 3D scene. During rendering, the computer “draws” the 3D objects to the screen and/or to a file, usually a bitmap. Real-time rendering occurs when viewing objects in interactive viewports, or in games or simulations. Non real-time production rendering draws a much higher quality image, at the expense of speed. A production rendering may take anywhere from a few seconds per frame to an indefinite period of time... sometimes hours or even days for a single frame.
Computer Vision: Discipline that studies the techniques and algorithms for constructing devices or systems capable of acquiring information from images.
Image-Based Rendering: It is a method for generating “full-parallax” still or video images from photographed and live-action images taken from multiple viewpoints without the need for modelling (i.e., a process that uses geometrical configurations to draw up shapes of objects).
Rendering Volumetric: Particular rendering technique that allows direct calculation of the volume of the objects to be rendered, without the intermediate step to a representation of surface. The standard method used for this type of rendering is Ray-Tracing. Its results are very realistic, but computationally much heavier compared to other techniques of global illumination.
Digital Archives: Archives of documents memorized with informatics procedures. They contain a large amount of data on one or more topics related to each other whose management (i.e. insert, editing, and research data) is carried out by means of special software. Useful for the understanding, preservation and dissemination of data relating to the cultural heritage.
Numeric Model: Synthetization of survey datum in which is registered every single information acquired, be it metric or chromatic. It is also known as “point clouds”. The term refers to a model whose shape is described by points characterized by their spatial coordinates x,y,z located in a Cartesian reference system.
Photomodelling: Methodology for non-contact survey that refers exclusively to the use of photographs to conduct the three-dimensional reconstruction of real objects. It is a work ambiance which provides the possibility of three-dimensional re-creation of buildings by globally and coherently integrating the stages of surveying, modelling and representation. This is achieved by extracting directly from photographs all the information necessary in each of these stages: coordinates, distances, characteristic points for the two dimensional restitution of planes and perspectives; apices and profiles for three dimensional reconstructing of elements; textures for visually enriching created volumes. The development of a three-dimensional model passes through 3 closely related phases: acquisition of the spatial coordinates (the points present on an image are associated with homologous points present on different images of the same scene, returning the scene within a single spatial reference), three-dimensional reconstruction of the geometry, restitution of visual appearance (the geometric nature of the scene is enhanced by the attributes that allow to describe aspects of the surface: the 3D model are associated with the texture acquired at the time of the shooting).
MeSH: A 3D object composed of triangular facets. A mesh object has no true curvature, geometrically is composed of a set of vertices joined by many small flat surfaces, usually triangular or square, one adjacent to another. The appearance of curvature is achieved by increasing the number of faces (level of detail), and by edge smoothing during render time.
Textured Model: Model obtained with techniques based on the digital development of images, uses texture mapping to define in detail the formal aspect and the state of conservation of the organism, making a continuous exchange between the represented object and the observer, a two-way relationship between iconic representation and virtualization of the surveyed object. Textured models are suitable for general users - since they closely resemble reality - and for expert users, because the photo is linked by the points surveyed by laser scanner.
3d Models: Communication tool for the understanding and support of the object based on three-dimensional representation of a real object in a virtual space. Depending on the objective for which they are created, two types of model can be made: figurative 3d models and 3d models for scientific purposes. The figurative 3d models are aimed at achieving a convincing documentation of reality with a purely informative purpose; they are characterized by the identification of the geometry that describes the object represented by the clear and readability of its component parts. The 3d models for scientific purposes have a higher dimensional and formal correspondence with the real object and are characterized by high accuracy and reliability of the dimensions represented. In the first case, any survey's methodology is able to obtain data necessary for a correct realization of the final model, while in the second one needs to have qualitative accurate and highly precise survey data.
Scientificity: Is a modality with which science proceeds to cognize reality. It must be objective, reliable, verifiable and sharable. It consists – on the one hand – in acquiring experiences by experimental observation, and on the other – in formulating hypotheses and theories whose efficacy is tested by experimenting.
3d Surveying: Knowledge system which includes the steps of acquisition data /do not understand/ with the aim to obtain the maximum objectivity. The techniques for the 3d survey are linked to non-contact survey, which have the aim to return the representation of a real object in a virtual three-dimensional space. Nowadays the methods of highly specialized 3d survey, as 3d laser scanner short range and long range, comparing with those expeditious one, as the photomodelling, image-based methodology characterized by greater speed in the phases of acquisition and elaboration of data.
Laser Scanning: Is a method of massive automatic capturing of 3d points at fixed intervals. The scanner is used to create a “point cloud” of the object's surface. These points can then be applied to extrapolate the surface of complex object. The devices emit an electromagnetic pulse (laser) and receive the signal reflected by the surface. This way allows one to measure the time lapsed between emitting the beam and its return to the instrument, and thus the distance between the instrument and the scanned point. Laser scanners are capable of capturing a large amount of data in a short time generating a very large set of points distributed on the scanned object with a spacing that depends on the degree of accuracy to be achieved. Each point is characterized by five types of information: the three numerical values of x,y,z coordinates; the RGB value used to map the color information in the photograph taken by the scanner to the point; the reflectance, another RGB value which corresponds to the quantity of energy emitted by the instrument that returns after striking the scanner surface.
Texture Map: A bitmap or algorithm used to create surface color on an object. Also known as a diffuse color map. Ironically, texture maps do not simulate roughness. Bump and normal maps produce the illusion of roughness.
Structure From Motion (SFM): Systems based on algorithms that allow one to get automatically three-dimensional data and color from a set of photographs representing the object of interest. These systems are able to automatically retrieve the position and orientation of the cameras used for a reconstruction, from a set of images in any order. It is studied in the fields of computer vision and visual perception.
Texture Mapping: Texture mapping is based on a stored bitmap consisting of texture pixels, or texels. It consists of wrapping a texture image onto an object to create a realistic representation of the object in 3D space. The object is represented by a set of polygons, usually triangles. The advantage is complexity reduction and rendering speed, because only one texel read is required for each pixel being written to the frame buffer. The disadvantage is the blocky image that results when the object moves.
NURBS: Non Uniform Rational B-Splines, an algorithmic method for the construction of curved surfaces and free-forms. A particular type of surface spline for making curved surface patches in modeling complex shapes. These types of splines are supported by many computer aided design (CAD) systems used for modelling.
2d Models: Procedure for the communication of knowledge of the built environment and multidimensional elements based on schematization that leads to a reduction in the number of dimensions from three to two and that materialize on two-dimensional supports. Drawings that are conventionally represented elements of architectural and archaeological nature at different scales of representation, consisting of plans, elevations and sections. In order to communicate all the knowledge acquired about the surveyed object, two types of models can be made: geometric models and textured models. The first typology of model is characterized by the geometrization of the elements represented and show the morphology and the spatiality of the object; the other one makes explicit the actual configuration of the elements and shows graphic characterization indicating the quality of the surfaces and their state of preservation.
Virtual Reality: Tool which allows to view - three-dimensionally with high-resolution - the environments and objects and interact with them in real time by providing a sense of immersion and presence in the built environment. Intended to replace the real world with a completely synthetic one, isolated from reality, in which the user is immersed and with which he can interact.
Graphics Workstation: Generic term for a high-performance desktop computer that is used in the context of CAD drawings, 3D modelling or video editing. Usually includes one or more 64-bit processors, ultra-fast drives connected in RAID, one or more tabs graphs of high performance such as NVIDIA Quadro.
Augmented Reality: Representation system that integrates computer generated images with the vision of the real world. It doesn't isolate the user from the real world, but rather complements him by virtual objects generated by computer, in a world that is made up of real and virtual objects. It increases perception and user interaction with the environment by providing visual information that the user could not directly detect with his own senses. The virtual world is virtually enriched with additional textual and graphic information, synchronized and generated by the computer. The objective is to increase the visual perception of physical space with images of virtual space in which the user can move freely in the scene, with the ability to interact with it.
Modelling: Construction of geometric objects in 3D scenes. Models describe the forms of objects, but not their material properties or the way they move.