3D digital recording

Complete digital recording is a multi-dimensional process. It depends highly on the nature of the subject of recording as well as the purpose of its recording. The whole process involves the three-dimensional digitization, digital data processing and storage, archival and management, representation and reproduction.

Link

Citation

Favorite

Collect this page

Introduction

Complete digital recording is a multidimensional process. It does not only address the problem of three-dimensional (3D) digitization but involves all the aspects of digital content management, representation and reproduction. It addresses issues affecting the whole life cycle of the digital content. Five main processes can be identified in digital recording, as shown in Fig. 1. All there processes have their own needs for advanced algorithms, new hardware and more sophisticated software implementations.

Fig.1. Main processes of 3D digital recording

3D Digitization

3D digitization is the first step of the overall process. It consists of multiple sub-processes and exhibits variations in accordance with specific application requirements. Due to the complexity of the digitization needs that emerge from the objects themselves, there is a plethora of methods and technologies. The target of every such technique is to address successfully a particular type of objects or class of objects, or to fulfill particular needs and demands of a specific digital recording project (i.e. complete recording for archival, digitization for presentation, digitization for commercial exploitation).

The plethora of available 3D digitization systems is the result of three main factors that influence the suitability and the applicability of a method:

1. Complexity in size and shape

2. Morphological complexity (level of detail)

3. Diversity of raw materials

Fig.2. Complexity in size and shape

Fig.3. Morphological complexity (level of detail)

Fig.4. Diversity of raw materials

There are techniques with satisfying results for microscopic objects, other for small, medium and large objects and other for large scale objects, open spaces and urban areas. Additionally, there are different techniques for ceramic, metallic or glass objects. It should be noted that techniques with satisfying results for one kind of objects are usually non-applicable to other. An extensive study on the available methods for 3D digitization produced the construction of the Nine-Criteria-Table, which summarizes the possible parameters for choosing a 3D digitization system.

TABLE I. The Nine-Criteria-Table for choosing a 3D digitization system

No	Criterion
1.	Cost
2.	Material of digitization subject
3.	Size of digitization subject
4.	Portability of equipment
5.	Accuracy of the system
6.	Texture acquisition
7.	Productivity of the technique
8.	Skill requirements
9.	Compliance of produced data with standards

3D digitization is a complex process that consists mainly of three phases:

1. Preparation, during which certain preliminary activities take place that involve the decision about the technique and methodology to be adopted as well as the place of digitization, security planning issues, etc

2. Digital recording, which is the main digitization process according to the plan from phase 1

3. Data processing, which involves the modeling of the digitized object through the unification of partial scans, geometric data processing, texture data processing, texture mapping, etc

Digitization of objects

Laser scanning techniques

Laser scanning techniques are based on a system with a laser source and an optical detector. The laser source emits light in the form of a line or a pattern on the surface of the objects and the optical detector (usually a digital camera) detects this line or pattern on the objects. By applying the well known triangulation principle the system is able to deduce the distance from the objects and to extract their geometry. The advantage of using laser sources is that laser light is very bright and highly focused for long distances. As a result the emitted pattern can be always focused on the surface of the objects.

One of the most significant advantages of laser scanners is their high accuracy in geometry measurements. On the other hand, it should be noted that in many such systems, geometry can be extracted without any texture information. Additionally, special attention should be paid for surfaces with specific properties, such as reflectance and transparency. One other important aspect is the high cost of such devices, which renders this method useful to specific-only applications. Finally, the productivity of the method, as well as the portability, depends upon the used system and can vary significantly [1]-[5].

Shape from structured light

This method is based on projecting a specific pattern on the surface of the objects and trying to extract geometry information from the deformations of this pattern. This method is also based on triangulation but does not need to use specific laser sources. In many cases this method is confused with the laser scanning methods and there are commercial systems that can not be absolutely categorized to the one or the other method.

The method works by projecting a specific predefined light pattern that covers the whole (or part of the) surface of the objects. This scene is then captured by a typical digital image detector and processed in order to deduce the geometry from the deformations of the pattern in the digital image. These patterns can be simple multiple fringes of different colors or complex patterns with curves, either time or space coded.

This method is accompanied by texture acquisition and can lead to very impressive results in terms of accuracy and productivity. The systems are usually portable and easy to use. A lot of work is still being done to develop even more the resolution of the method, which is one of the main fields of research in 3D scanning today [6]-[33].

Shape from silhouette

This technique is based on multiple photographic capturing of the object from different viewing angles, and deducing the geometry from the object’s silhouettes. This is, actually, an old idea originating back to 1960 when Francois Villeme discovered a method called photo-sculpturing: 24 photographs covering the surface of the object are taken and are projected onto clay. This method regained interest about 100 years later with the advent of computers.

Recent improvements of this method use texture information to correct or enhance geometry with very interesting results in terms of the final recorded geometry.

Shape from silhouette is an automated process with high productivity and relatively low cost. As of this it is very popular. It can capture both geometry and texture. It is portable and easy to use. The main disadvantage is the medium-to-low resolution in geometry measurements [34]-[44].

Shape from stereo

Main goal of this method is the extrapolation of as much geometry information as possible from only a pair of photographs taken from known angles and relative positions, simulating the human visual system.

Stereo-photography has a significant application in robotic and computer vision. It is based on taking pairs of photographs from slightly different angles. When certain parts of the object in the scene are visible to both photographs, specific algorithms from vision can be applied to extract the geometry of the object. The external as well as the internal parameters of the optical system are used for calibration. Calibration is critical in terms of achieving accurate measurements. The method can either be fully automated or man-operated. The final result is a depth map of the object in the scene, reflecting the distance of each recognized point on the surface of the object from the photographic sensor.

Advantages of this method are the ability to capture both geometry and texture, the low cost and the portability. Disadvantage of the method is its low resolution [45]-[47].

Shape from video

Shape from video is a variant of shape from stereo. Here the two photo cameras are replaced by a video camera that captures the object in a sequence of images from different views. A basic requirement for the application of this method is that the object is at complete rest and with no movable parts.

The algorithms that are being used are similar to the ones in shape from stereo and are sensitive to noise in the video sequence. A key point in the process is the identification of common points between different images and the registration of these points onto a virtual 3D scene.

The results are, sometimes, ambiguous due to the fact that there is no prior knowledge about the position of the camera or the objects. Advantages of this method can be considered the low cost, the portability and the ability to capture both texture and geometry. Significant disadvantage is the low resolution in capturing the geometry [48]-[49].

Shape from shading

Shading plays an important role in depth perception. Many researches have already tried to simulate the way the human visual system uses shading information to perceive the depth. This method requires the capturing of the object from one viewing angle. What should vary is the position of the light source, which would cause the shading to vary on the surface of the object. This way, special algorithms could deduce the geometry of the surface of the object by using multiple photos of different shading conditions.

The method is simple and has low cost. It captures both geometry and texture, with a minor disadvantage in capturing texture in shaded areas. It is portable but has the disadvantage of low accuracy. There are thoughts of using this method combined with other methods (such as shape from stereo) in order to enhance its accuracy [50].

Shape from texture

Texture can be a very significant source of information for the surface geometry. The calculation of 3D primitive shapes on a surface can be done if there exists some prior knowledge about the surface texture. It is well known that tha human visual system can easily identify surface geometry when the surface texture is homogeneous. Researchers tried to exploit this observation in order to simulate the process. So, the idea is to identify small structuring texture elements (texels) and to find their possible transformations in order to reproduce the whole surface of an object. These identified transformations are then used to extract the actual 3D surface geometry.

The method is, again photographic, simple and of low cost, but has limited applications (like capturing of fabric or human skin). It is portable and easy to use. On the other hand it has low accuracy [51].

Shape from photometry

Shape from photometry is a variant of shape from shading. Here the photos show the object from one viewing angle but varying lighting conditions. Additionally, the usage of reference objects (or, in some cases, reference lighting sources) in the scene is critical, since they are used as calibration objects. Calibrated lights can improve significantly the result of the method but can only be found in special laboratories, so in this case the method is not portable. In studies against laser techniques there have been reports that favor this method in terms of the produced data volume and the immunity to laser limitations. Generally it can be regarded as portable, and it is easy to use and of low cost. Main disadvantage is its current need for laboratory environment [52]-[54].

Shape from focus

Lately, a new possibility has attracted the interest of researchers: the possibility of exploiting the depth of field in a photo in order to deduce the 3D geometry of the scene. The method is recursive and is based on taking photos of an object while adjusting continuously the focus plane. By knowing the position of this focus plane (from the whole setup and the positioning of the system) we are able to map the focused pixels in an image on the correct position in the 3D depth map. The system, recursively, rebuilds the whole object geometry photo by photo.

Resolution, as well as accuracy is limited, but the results are, in general, “reliable”. One limitation comes from the fact that in order to take photos with so limited depth of field one might need very special lens and a major application is in the usage of microscopes. The cost is relatively high, but the method is simple and easy to apply [55]-[57].

Shape from shadow

Shape from shadow rebuilds the 3D model of an object by exploiting the deformation of the shadow of a known object which is projected onto the surface of the subject of digitization, when the light is moving. As obvious, this is a simple variant of the shape from structured light technique.

Main advantage of this method is the low cost and the limited demand for computing power. It can reconstruct geometry even in non visible parts on the object, under certain assumptions about the object (or any prior knowledge). For this method, one might even find open source code on the Internet. The method has low accuracy [58].

Contact systems

Very often we might find digitization systems that use lasers mounted on some sort of arm with high degree of freedom. This arm can either be operated manually or automatically, and through its internal positioning system carries geometry information to the controlling software. These arms are generally called Coordinate Measuring Machines - CMM. Apart from their combined usage with laser devices, they can be operated autonomously. A spike can be mounted on them and by maintaining continuous contact with the object to be measured, very accurate geometry information can be recorded.

The method is of high accuracy but very show. There is also the disadvantage of having to be in contact with the objects, which is sometimes inadmissible [59].

Digitization of monuments

Empiric

During an empiric recording of monuments, measurements of distances between characteristic points on the surface of the objects are taken manually . The definition of the coordinates is being done on an arbitrary coordinate system on a plane surface of the object.

The method is simple and productive, portable and of low cost. On the other hand it is of low accuracy and demanding in terms of time of physical presence near the object. It can be successfully applied when an object has simple shape, or there is a need for recording a sectional plan or sections of interiors [60].

Topographic

The topographic method implements an 3D orthogonal coordinate system by using complicated and high accuracy measuring devices. Mainly, this method using a Geodesic Station, a system for measuring angles and distances of characteristic points on the surface of the objects, which are further transformed to coordinates in reference to the initial orthogonal coordinate system.

Main advantage of the method is its high accuracy and objectivity of the measurements. It is reliable and it is easy to process its results. A disadvantage is the need of long physical presence near the object, but it is one of the only methods to be used under difficult conditions, such as complex shape and difficult access. It is referred as ideal for producing high accuracy models of scale 1:50 or smaller [60].

Laser scanning techniques

Laser scanners can actually be considered as advanced geodesic stations and can be used to measure topographic quantities. They can measure the direction of a fictional optical line joining the characteristic points on the surface of the objects to a reference point on the measuring device. Additionally these scanners can estimate their distance from these points. By applying the known triangulation principle they produce Cartesian coordinates automatically.

Main advantage is the high accuracy and productivity, as well as the large volume of produced measurement data. It is reliable and objective. On the other hand it is a method of high cost and difficulties in portability and autonomy. It can be applied on almost every large scale digitization, but can experience interference from very bright light [61].

Photogrammetry

Common digital photos can be used, under suitable conditions, for measurements that can be of the accuracy obtained by the topographic methods. By applying orientation processes and transformations of digital photogrammetry it is possible to deduce 2D or 3D coordinates from one or two photos.

The method is objective and reliable and can be aided by CAD software. It is relatively simple and has low cost. On the other hand it has to be combined with topographical or empirical measurements and the final outcome is a function of the time spent. It can be used for complex objects with high surface detail, but since it is based on photos, there is a need for adequate space. It is also useful when direct access or contact to the object is prohibited. It can be used to record the condition of the object in time. When combined with accurate measurements it can produce models of high accuracy for scales of 1:100 and even higher [60],[62],[63].

3D Digital Data Storage

Data produced by the process of 3D digitisation are usually large files, the size of which depends on both the size of the digitised subject and the resolution of digitisation [64]-[69].

The raw material delivered by the 3D digitisation hardware (3D scanner) most of the times is available in a file format which is recognised only by the software used for the acquisition of that data and it is initially stored on the local hard drive of the digitisation workstation. Afterwards, that raw material must be collected and adequately processed, in order to produce the anticipated for each application result. More or less, the procedures involved after the end of 3D scanning are:

1. Storage of raw material for archiving purposes. Accompanying metadata information is mandatory.

2. Construction of a unified form from the raw material, representative of the digitised subject. Metadata must not be excluded from the storage of that form.

3. Conversion, of the outcome from the 2^nd stage, to a file format more common to the 3D computer graphics industry. Storage including metadata information is compulsory for the products of this stage also.

4. Further conversion to file formats that are more adequate for the needs of a specific application.

Depending on application, metadata information can be determined by the following three basic categories [69]-[71].

1. Descriptive information on the subject, provided by the user.

2. Information for the administration of the data (i.e. version control etc)

3. Algorithmic description of the data, for the purpose of content based search and retrieval of the information. Therefore the results from a content based database query are not affected by the error prone user based description of the data.

For the selection of the device which will be used for the storage of the huge amount of digital information, produced by the 3D digitisation process, we must take into consideration the following characteristics:

1. Data access time.

2. Data transfer rate, from the storage device to the computer memory and vice versa.

3. Multi user access capabilities.

4. Digital storage capacity

5. Utilisation frequency.

6. Data retention lifetime of the storage medium.

7. Required environmental conditions for the storage and operation of the storage device / medium.

8. Cost per digital storage unit (Megabyte / Gigabyte).

In order to conserve a digital collection as much as possible, we have to keep up with the rules prescribed by the manufacturer of the selected storage medium. Appropriate handling and storing of the storage medium is vital for its lifetime and consequently for the survival of the stored information too. In order to prolong data’s lifetime even more, periodic storage medium check ups and precautionary copying of their data to a newer medium of the same type is a good practice. However, due to the frenzied evolution of computer hardware, both storage devices and their storage media are rendered obsolete before reaching even the half of their lifespan. Thus, data migration to a tested modern storage solution is the best practice. The following table depicts some of the characteristics of today’s storage media.

TABLE II. Characteristics of today’s storage media

Criteria	Score ( max is better)
	5	4	3	2	1
Capacity	Tape	HDD(**)	Rem.	DVD, Μ/Ο	CD, FDD, Flash
Data transfer rate	HDD	Rem.	DVD, CD, Flash	Tape, Μ/Ο	FDD
Data access time	Flash	HDD	Rem.	DVD, CD, FDD, M/O	Tape
Storage device cost	Tape, Μ/Ο	Rem.	HDD	DVD, CD	FDD, Flash
Cost / MB (Storage Unit)	DVD	Tape	HDD	Rem., CD, M/O, FDD	Flash
Re-writability	HDD, Rem.	M/O, Tape	Flash, FDD	CD/DVD-RW	CD/DVD-R
Reliability	M/O, HDD, Rem.	DVD, CD	Tape	Flash, FDD
Fields resistance	DVD, CD	Flash, M/O	HDD	Rem.	FDD, Tape
Portability	Flash	FDD, DVD, CD, Rem.	HDD, M/O	Tape
Lifespan	HDD, DVD, CD, M/O	Rem, Tape		Flash	FDD
*Sensitivity ()**	Flash	M/O	Tape, FDD	HDD, Rem.	CD, DVD

(*) Storage medium sensitivity to frequent and careless use

(**) Table Legend:

Tape	Magnetic tape
HDD	Hard disc drive
Rem	Removable hard magnetic disc
M/O	Magneto-Optical
FDD	Removable floppy magnetic disc
DVD	Digital Versatile Disc
CD	Compact Disc
Flash	Solid state flash memory

3D Digital Data Reproduction

The reproduction of a 3D digitised subject is feasible via two different ways:

1. Digital Reproduction of data. The most common devices for making multiple copies of digital subjects, in order to make that information available to computer users without access to broadband internet connection, are the optical disc (CD / DVD) multiplication devices.

TABLE III. Digital data reproduction options

Type	Storage medium	Max recording speed	Operation
DVD Copy Tower	CD / DVD	8 X DVD 52 X CD	Manual disc interchange
CD Copy Tower	CD	52 X CD	Manual disc interchange
Robotic CD / DVD Copier	CD / DVD	8 X DVD 52 X CD	Automatic disc interchange for (50 – 500) discs

2. Physical reproduction of data. 3D printing, or solid imaging, is another way for reproducing the information gathered from the process of 3D digitisation. 3D printing is the process of creating a tangible copy of intangible digital data, using a special device which is able to construct the material representation of the 3D dataset.

TABLE IV. Physical reproduction options

Type	Material	Printed object dimensions (cm)	Application
Matter deposition	plaster, starch, plastic, metal powder	20Χ20Χ25 to 50Χ60Χ40	Mould and prototype construction
Stereo lithography	Photopolymer	25Χ25Χ25 to 50Χ60Χ50	Mould and prototype construction
3D Milling	Foam, wood, plastic, metal, stone	25Χ25Χ25 to 500Χ500Χ500	Mould and prototype construction

3D Digital Data Visualization

For the electronic presentation of 3D digitised subjects we can choose from a wide variety of electronic display equipment, depending on the specific requirements of each application. The parameters that we must consider for the selection of the most adequate imaging device are [72]-[74]:

The number of people that are watching simultaneously the presentation. That determines the size of the display device.
Stereoscopic presentation of the 3D subject, in order to set the 3^rd dimension perceivable by the user.
User interaction degree. For example, whether the user will be able to wander freely in the virtual 3D world or be restricted in a certain course. That parameter of presentation determines the type of hardware which will provide the images to the electronic display device. In the case of the predetermined course, that device could be a simple video player which costs less than 100 €. However, in case of free wandering in the 3D world, the image supplying device could vary from a simple personal computer, with cost of a couple of hundreds €, to a visualisation workstation that costs a lot of thousands €.

TABLE V. 2D display options

Type	Max display size (inches)	Max image resolution (pixels)	Depth perception technique	Lighting requirements	Image source	Portability
Digital cinema projector	Cinema theater size	2048 Χ 1080	Combining two polarised projectors. LCD shutter glasses. Blue red glasses	Moderate	DVI, VGA, SVideo, Composite, RGB-BNC	Moderate or Bad
Data/ Home cinema projector	30 - 220	1280 Χ 1024	Combining two polarised projectors. LCD shutter glasses. Blue red glasses	Low light condition	DVI, VGA, SVideo, Composite.	Good
TFT	14 - 57	1920 Χ 1080	Blue red glasses	Moderate	DVI, VGA, SVideo, Composite.	Good
Plasma	17 – 65	1365 X 768	Blue red glasses	Moderate	DVI, VGA, SVideo, Composite, RGB-BNC	Good
CRT	15 - 34	2048 X 1536	LCD shutter glasses. Blue red glasses	Very few	VGA, SVideo, Composite, RGB-BNC	Moderate

A lot of companies are using projectors in various ways in order to construct special display devices, such as walls, caves, domes etc. These display devices are known as immersive displays and their cost is usually huge. Depending on the application, the cost usually starts form a couple of thousands € and can easily surpass 100,000 €.

Except the common 2 dimensional display systems, there are electronic devices that are able to display true 3D information, without demanding the use of special accessories, like glasses, in order to achieve that. These devices are known as active 3D displays and are available in three different types:

1. Flat panel displays (Plasma and TFT), which provide the sense of depth via special lenticular filters in front of the display.

2. 3D Volumetric displays, where the third dimension is produced by projecting the appropriate sequence of images on a very fast moving semitransparent surfaces.

3. Head mounted displays (HMD). These are wearable devices which provide a small display screen in front of each eye. Those displays are independent, so by feeding them with the appropriate images it is possible to achieve the required effect of true stereoscopic vision.

TABLE VI. 3D display options

Type	Max display size (inches)	Max image resolution (pixels)	Lighting requirements	Image source	Portability
Flat panel (TFT or Plasma)	50	3840 Χ 2400	Moderate	DVI, VGA	Good
3D Volumetric Display	20	768 X 768 X 198	Low lighting conditions	Through special hardware	Moderate
HMD	A mini display in front of each eye.	1600 Χ 1200	None	DVI, VGA, or Through special hardware	Very good

References

[1] Curless, B.; Levoy, M., “Better optical triangulation through spacetime analysis”, Proceedings of Fifth International Conference on Computer Vision, 20-23 June 1995, pp. 987 – 994.

[2] Rioux, M., 1994. “Digital 3-D Imaging: Theory and Applications”, SPIE Proceedings, Videometrics III,International Symposium on Photonic and Sensors and Controls for Commercial Applications, Boston, 2650, pp. 2-15.

[3] J-Angelo Beraldin, Francois Blais, Luc Cournoyer, Guy Godin and Marc Rioux, “Active 3D sensing”, SCUOLA NORMALE SUPERIORE PISA, Centro di Ricerche Informatiche per i Beni Culturali, 2000.

[4] Josep Forest, Joaquim Salvi, Enric Cabruja and Carles Pous, “Laser stripe peak detector for 3D Scanners. A FIR filter approach”, 17th International Conference on Pattern Recognition, ICPR 2004, Cambridge, UK, 23-26 August 2004.

[5] Gabriele Guidi, Senior Member, IEEE, J.-Angelo Beraldin and Carlo Atzeni, “High-Accuracy 3D modeling of Cultural Heritage: The Digitizing of Donatello’s Maddalena”, IEEE Transactions on Image processing, Vol.13, No.3, March 2004.

[6] Salvi, J., Pages, J., Batlle, J., “Pattern codification strategies in structured light systems”, PR (37), No. 4, April 2004, pp. 827-849.

[7] J. L. Posdamer, M. D. Altschuler, “Surface measurement by space-encoded projected beam systems”, Computer Graphics and Image Processing, 18 (1) (1982) 1–17.

[8] S. Inokuchi, K. Sato, F. Matsuda, “Range imaging system for 3-D object recognition”, in Proceedings of the International Conference on Pattern Recognition, 1984, pp. 806–808.

[9] M. Trobina, “Error model of a coded-light range sensor”, Technical report, Communication Technology Laboratory, ETH Zentrum, Zurich (1995).

[10] C. Rocchini, P. Cignoni, C. Montani, P. Pingi, R. Scopigno, “A low cost 3D scanner based on structured light”, in A. Chalmers, T.-M. Rhyne (Eds.), EG 2001 Proceedings, Vol. 20(3), Blackwell Publishing, 2001, pp. 299–308.

[11] D. Caspi, N. Kiryati, J. Shamir, “Range imaging with adaptive color structured light”, Pattern analysis and machine intelligence, 20 (5) (1998) 470–480.

[12] Gühring, “Dense 3-d surface acquisition by structured light using off-the-shelf components”, Videometrics and Optical Methods for 3D Shape Measurement, 4309 (2001) 220–231.

[13] O. Hall-Holt, S. Rusinkiewicz, “Stripe boundary codes for real-time structured-light range scanning of moving objects”, in The 8th IEEE International Conference on Computer Vision, 2001, pp. II: 359–366.

[14] M. Maruyama, S. Abe, “Range sensing by projecting multiple slits with random cuts”, Pattern Analysis and Machine Intelligence, 15 (6) (1993) 647–651.

[15] N. G. Durdle, J. Thayyoor, V. J. Raso, “An improved structured light technique for surface reconstruction of the human trunk”, in IEEE Canadian Conference on Electrical and Computer Engineering, Vol. 2, 1998, pp. 874–877.

[16] J. Salvi, J. Batlle, E. Mouaddib, “A robust-coded pattern projection for dynamic 3d scene measurement”, International Journal of Pattern Recognition Letters (19) (1998) 1055–1065.

[17] E. M. Petriu, Z. Sakr, S. H. J. W., A. Moica, “Object recognition using pseudo-random color encoded structured light”, in Proceedings of the 17th IEEE Instrumentation and Measurement technology Conference, Vol. 3, 2000, pp.1237–1241.

[18] P. Lavoie, D. Ionescu, E. Petriu, “A high precision 3D object reconstruction method using a color coded grid and nurbs”, in Proceedings of the International Conference on Image Analysis and Processing, Venice, Italy, 1999, pp. 370–375.

[19] L. Zhang, B. Curless, S. M. Seitz, “Rapid shape acquisition using color structured light and multi-pass dynamic programming”, in Int. Symposium on 3D Data Processing Visualization and Transmission, Padova, Italy, 2002.

[20] C. Chen, Y. Hung, C. Chiang, J. Wu, “Range data acquisition using color structured lighting and stereovision”, Image and Vision Computing, 15 (1997) 445–456.

[21] E. M. Petriu, T. Bieseman, N. Trif, W. S. McMath, S. K. Yeung, “Visual object recognition using pseudo-random grid encoding”, in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems, 1992, pp.1617–1624.

[22] H. J.W. Spoelder, F. M. Vos, E. M. Petriu, F. C. A. Groen, “Some aspects of pseudo random binary array-based surface characterization”, IEEE Transactions on instrumentation and measurement 49 (6) (2000) 1331–1336.

[23] P. Griffin, L. Narasimhan, S. Yee, “Generation of uniquely encoded light patterns for range data acquisition”, Pattern Recognition 25 (6) (1992) 609–616.

[24] R. A. Morano, C. Ozturk, R. Conn, S. Dubin, S. Zietz, J. Nissanov, “Structured light using pseudorandom codes”, Pattern Analysis and Machine Intelligence 20 (3) (1998) 322–327.

[25] B. Carrihill, R. Hummel, “Experiments with the intensity ratio depth sensor”, in Computer Vision, Graphics and Image Processing, Vol. 32, Academic Press, 1985, pp. 337–358.

[26] T. Miyasaka, K. Kuroda, M. Hirose, K. Araki, “High speed 3-D measurement system using incoherent light source for human performance analysis”, in Proceedings of the 19th Congress of The International Society for Photogrammetry and Remote Sensing, The Netherlands, Amsterdam, 2000, pp. 65–69.

[27] G. Chazan, N. Kiryati, “Pyramidal intensity-ratio depth sensor”, Technical report 121, Center for Communication and Information Technologies, Department of Electrical Engineering, Technion, Haifa, Israel (October 1995).

[28] J. Tajima, M. Iwakawa, “3-D data acquisition by rainbow range finder”, in International Conference on Pattern Recognition, 1990, pp. 309–313.

[29] T. Sato, “Multispectral pattern projection range finder”, in Proceedings of the Conference on Three-Dimensional Image Capturer and Applications II, Vol. 3640, SPIE, San Jose, California, 1999, pp. 28–37.

[30] C. Wust, D.W. Capson, “Surface profile measurement using color fringe projection”, Machine Vision and Applications 4 (1991) 193–203.

[31] E. Horn, N. Kiryati, “Toward optimal structured light patterns”, Image and Vision Computing 17 (2) (1999) 87-97.

[32] J. Salvi, J. Pagès, J. Batlle, “Pattern codification strategies in structured light systems”, Pattern Recognition. Volume 37, Issue 4, April 2004, Pages 827-849.

[33] C. Rocchini, P. Cignoni, C. Montani, P. Pingi, R. Scopigno, “A low cost 3D scanner based on structured light”, in A. Chalmers, T.-M. Rhyne (Eds.), EG 2001 Proceedings, Vol. 20(3), Blackwell Publishing, 2001, pp. 299–308.

[34] Z. Zhang, “Modeling Geometric Structure and Illumination Variation of a Scene from Real Images”, In Proc. International Conference on Computer Vision (ICCV’98), Bombay, India, January 4–7, 1998.

[35] Tosovic S., Sablatnig R., Kampel M., “On combining shape from silhouette and shape from structured light”, in: H. Wildenauer and W. Kropatsch, (Eds.), Proc. of 7th Computer Vision Winter Workshop, pp. 108-118, 2002.

[36] A. Laurentini, “The Visual Hull Concept for Silhouette-Based Image Understanding”, IEEE Transactions on Pattern Analysis and Machine Intelligence, v.16 n.2, p.150-162, February 1994

[37] Sablatnig R., Tosovic S., Kampel M., “Next View Planning for Shape from silhouette”, in: Drbohlav O., (Ed.), Proc. of 8th Computer Vision Winter Workshop 2003, Valtice, Czech Republic, pp.77-82, 2003.

[38] Adam Baumberg, Alex Lyons, Richard Taylor, “3D S.O.M. – A commercial software solution to 3D scanning”, Vision, Video, and Graphics (2003), The Eurographics Association 2003. Eurographics Partner Event Vision, Video, and Graphics 2003. Bath UK.

[39] Potmesil, M. “Generating Octree Models of 3D Objects from their Silhouettes in a Sequence of Images”, CVGIP 40, 1987, pp. 1-29.

[40] Noborio et al., “Construction of the octree approximating three-dimensional objects by using multiple views”, IEEE Trans. on PAMI, Vol.10, pp.769-782, 1988.

[41] N. Ahuja and J. Veenstra, “Generating Octrees from Object Silhouettes in Orthographic Views”, IEEE Trans. Pattern Analysis and Machine Intelligence, February 1989, pp. 137-149.

[42] Y. Matsumoto, H. Terasaki, K.Sugimoto and T. Arakawa, “A Portable Three-dimensional Digitizer”, IEEE 1997, 3-D Digital Imaging and Modelling, 1997.

[43] Hendrik P.A. Lensch, Wolfgang Heidrich, Hans-Peter Seidel, “A silhouette-Based algorithm for texture registration and stitching”, 2001, Elsevier Science (USA), 1524-0703/01.

[44] A. Laurentini, “The Visual Hull Concept for Silhouette-Based Image Understanding”, IEEE Transactions on Pattern Analysis and Machine Intelligence, v.16 n.2, p.150-162, February 1994.

[45] D. Scharstein and R. Szeliski. “A Taxonomy and Evaluation of Dense Two-Frame Stereo Correspondence Algorithms”, IJCV, 2002.

[46] M. Bertozzi, A. Broggi, G.Conte and A. Fascioli, “Stereo-Vision System performance analysis”, Enabling Technologies for the PRASSI Autonomous Robot, pages 68-73. ENEA, Rome, Italy, January 2002, ISBN 8882860248.

[47] A. Streilein, F. a.van den Heuvel, “Potential and limitation for the 3D documentation of cultural heritage from a single image”, CIPA International Symposium 1999, October 3-6, 1999, Recife/Olinda - PE – Brazil.

[48] A. Chiuso, H. Jin, P. Favaro and S. Soatto (2000). “MFm: 3-D Motion and Structure from 2-D Motion Causally Integrated Over Time: Implementation”, In Computer Vision - ECCV 2000, D. Vernon ed., Lect. Notes in Computer Science 1843, pp. 734-750.

[49] L. Van Gool, Filip Defoort, Marc Pollefeys, Reinhard Koch, Marc Proesmans, Maarten Vergauwen, “Special Lecture: 3D Modeling for Communications”, p. 482, Computer Graphics International 1998,June 22 - 26, 1998,Hannover, Germany.

[50] Ruo Zhang, Ping-Sing Tsai, James Edwin Cryer, Mubarak Shah, “Shape from Shading: A Survey”, IEEE Transactions on Pattern Analysis and Machine Intelligence, August 1999 (Vol. 21, No. 8).

[51] D.A. Forsyth, “Shape from texture without boundaries”, Proceedings of the 7th European Conference on Computer Vision-Part III, pp. 225 - 239, 2002, ISBN:3-540-43746-0.

[52] Chia-Yen Chen, Reinhand Klette, Chi-Fa Chen, “3D Reconstruction Using Shape from Photometric Stereo and Contours”, Image and Vision Computing, Palmers ton North, November 2003.

[53] R. Basri and D. Jacobs, “Photometric Stereo with General, Unknown Lighting”, IEEE Conference on Computer Vision and Pattern Recognition 2001.

[54] Aaron Hertzmann, Steven M. Seitz, “Shape and Materials by Example: A Photometric Stereo Approach”, Proceedings of CVPR 2003, IEEE Computer Society Conference on. Computer Vision and Pattern Recognition.

[55] P.J. Besl. Active, “Optical range imaging sensors”, Machine Vision and Applications, 1.127-152, 1988.

[56] Yoav. Y. Schechner, “Depth from Defocus vs. Stereo: How different really are they?”, International Journal of Computer Vision 89 pp. 141-162 (2000).

[57] Paolo Favaro, “Shape from Focus/Defocus”, Washington University Department of Electrical Engineering Electronic Signals and Systems Research Lab, June 25th, 2002, http://homepages.inf.ed.ac.uk/rbf/CVonline/LOCAL_COPIES/FAVARO1/dfdtutorial.html

[58] Jean-Yves Bouguet, Pietro Perona, “3D Photography on your desk”, in Proc. Of the Int. Conf. On Computer Vision, Bombay, India, January 1998.

[59] M. Nashman. , T. Hong. , W. Rippey. , M. Herman, “An Integrated Vision Touch-Probe System for Dimensional Inspection Tasks”, Proceedings of the SME Applied Machine Vision `96 Conference, Cincinnati, OH, June 3-6, 1996.

[60] Livieratos,E., Emperic, “Toporgaphic or Photogrammetric recording? Answers to properly phrased questions”, Proceedings of the congress, “Terrestrial Photogrammetry and Geographic Information Systems for the documentation of the National Cultural Heritage”, Thessaloniki, 1992.

[61] Boehler, W., A. Marbs, “3D Scanning Instruments”, Proceedings Of The CIPA WG 6 International Workshop, “On Scanning For Cultural Heritage Recording”, Corfu, 2002.

[62] Hanke K., Grusenmeyer, P., “Architectural Photogrammetry: Basic theory, Procedures, Tools”, Tutorial of Architectural Photogrammetry, Corfu, September 2002.

[63] Tsioukas, V., Patias, P., “Low Cost 3D Visualization and Measuring ‘Tool’ in the Service of Archaeological Excavations”, Computer Applications and Quantitative Methods in Archaeology Conference, “The Digital Heritage of Archaeology”, Heraklion, 2002.

[64] Albert S. Hoagland, “Information Storage: Yesterday, Today and Tomorrow” Magnetic Disk Heritage Center (MDHC) 6/6/2002.

[65] P.G.Hulme, Xyratex UK, “Mass storage the options”, International Broadcasting Convention, 12-16 September 1996, Conference publication No. 428, IEE 1996.

[66] Ann Marie Przybyla, Geof Huth, “Preparing for the Worst: Managing Records Disasters”, The University of the State of New York, 2004.

[67] T. Schwarz, “Magnetic Tape as the Mass Storage Medium”, IEEE-NASA MSS Conference 2000. http://romulus.gsfc.nasa.gov/msst/conf2000/VG/C04VG.PDF

[68] M. Leonhardt, “Future Directions for Recording Technologies”, THIC Meeting at StorageTek, July 2003. http://www.thic.org/pdf/Jul03/stk.mleonhardt.030722.pdf.

[69] Helen Heslop, Simon Davis, Andrew Wilson, “An Approach to the Preservation of Digital Records”, National Archives of Australia, December 2002.

[70] “Cedars Guide to Digital Collection Management” The Cedars Project, March 2002, http://www.leeds.ac.uk/cedars/guideto/collmanagement/.

[71] B.S. Manjunath, Philippe Salembier, Tomas Sikora, “Intrduction to MPEG – 7”, John Wiley & Sons, Ltd.

[72] Larry J. Hornbeck, “Digital Light Processing—Introduction From cathode rays to digital micromirrors: A history of electronic projection display technology”, Texas Instruments TECHNICAL JOURNAL, July–September 1998.

[73] Gregg E. Favalora, Joshua Napoli, Deirdre M. Hall, Rick K. Dorval, Michael G. Giovinco, Michael J. Richmond, Won S. ChunActuality Systems, Inc. “100 Million-voxel volumetric display”, Society of Photo-Optical Instrumentation Engineers, 2002.

[74] C van Berkel et al, “Design & Apps of 3D-LCD”, Proc SID Euro-Display96, pp. 109-112, 1996.