The stereoscopy ( ancient Greek στερεός stereos , space / spatially fixed 'and σκοπέω skopeo consider') is the reproduction of images with a spatial impression of depth, which is not physically present. Colloquially, stereoscopy is incorrectly referred to as “ 3D ”, even though it is only two-dimensional images ( 2D ) that convey a spatial impression (“spatial image”). Normal two-dimensional images without impression of depth are called monoscopic (Greek: μονος, monos “one” → simple).
The principle is always based on the fact that humans, like all primates and most predators , look at their surroundings from two angles at the same time through their two eyes . This enables your brain to efficiently assign a distance to all objects being viewed and to gain a spatial image of its surroundings (“ spatial vision ”) without having to keep your head moving. The stereoscopy is therefore only concerned with bringing different two-dimensional images from two slightly different viewing angles into the left and right eyes.
There are various methods of doing this.
All other properties of a two-dimensional image, such as perspective distortion depending on an unnatural focal length of the lens , the color and, in particular, the restrictive location of the viewer, are retained. The last two properties of this spatial image method in particular cause the considerable difference to holography , which deals with the attempt to record and reproduce objects completely, i.e. three-dimensionally (in 3D).
When looking at close-up objects, binocular vision is an essential means of correctly estimating distances. With the right eye we see a close object projected onto a different part of the fundus than with the left eye, and this difference becomes more significant the closer the object gets. If we direct both eyes to one point, the two axes of the eyes form an angle that becomes larger the closer the object is. Close objects are seen a little more from one side with the right eye and a little more from the other side with the left eye. These two images, which can not exactly coincide due to the laterally disparate shift, but nevertheless lie within the so-called Panum area , are combined to form an overall spatial impression (spatial image), which is therefore essentially composed of two pieces of information: the different perspective Both eyes produce two different images and the curvature of the lens adapts to the distance of the object seen in order to produce a sharp image on the retina . The size of the viewing angle and the degree of accommodation provide a measure of the distance of the objects. The spatial resolution is therefore particularly high in the reach area. In addition, concealment and blurring effects as well as the perspective convey the spatial impression both binocularly and monocularly.
In the case of a stereo photo, only the information formed from the different angles is offered to the eyes. Since the eye tries to adapt the lens refractive power to the supposed distance, a sharp image on the retina only comes about after a certain delay (in the millisecond range). The contradiction between the supposed distance of the seen object and the actual curvature of the lens also causes dizziness or physical discomfort in some people after prolonged exposure (mismatch between vergence and curvature of the lens ).
The result of an unreal-appearing picture comes about when the stereo photo is presented sharply in all levels in order to achieve the spatial impression in full depth. In nature, however, only a certain area can be seen in focus ( depth of field of the eye). In order not to overwhelm the sense of sight, the manageable area can be deliberately limited when recording (see below: Lüscher-Winkel ).
The two required partial images are recorded simultaneously (synchronously) with a stereo camera that has two lenses at eye distance, also referred to as a natural base . Each individual image is referred to as a stereoscopic partial image , the image pair as a stereoscopic image . However, if the desired subject is still motifs ( still life , landscape ), the required partial images can also be recorded one after the other (metachronically) with a simple camera.
An enlargement or reduction of the base when recording enlarges or reduces the spatial impression when viewed. But even when recording with a natural basis, you have to take into account the different limits of the maximum tolerable deviation (deviation). It is Lüscher's merit to have pointed them out.
As early as the 4th century BC, the Greek mathematician Euclid dealt with spatial geometry ( stereometry ) in volumes 11–13 of his textbooks on mathematics . But he did not know that two eyes are necessary for a physiologically spatial visual impression.
In 1838 Sir Charles Wheatstone (1802–1875) published his first research on spatial vision. He calculated and drew pairs of stereo images and constructed an apparatus for viewing them in which the viewer's gaze was diverted to the partial images by mirrors. He called this device a stereoscope. Wheatstone achieved the union of the two partial images through his mirror stereoscope consisting of two mirrors inclined at right angles to each other , the planes of which are vertical. The observer looked into the left mirror with his left eye and into the right mirror with his right eye. Two sliding boards were attached to the side of the mirrors, bearing the inverted perspective drawings of an object. The rays emanating from the corresponding points in the two drawings were reflected by the mirrors in such a way that they appeared to come from a single point located behind the mirrors. So each eye saw the image belonging to it, and the observer received the spatial impression.
After Louis Daguerre announced the process of producing photographic images on layers of silver in 1839 at the Academy of Sciences in Paris , it made sense to use it to produce stereoscopic double images , which until then had only been available in drawn form.
In 1849 Sir David Brewster (1781–1868), Scottish physicist and private scholar, presented the first two-lens camera with which one could capture moving snapshots stereoscopically for the first time. Until then, the partial stereo images had to be exposed one after the other and the camera had to be moved between the two images at eye distance, which could lead to different image contents with moving subjects that did not allow a spatial impression.
In the same year, Brewster simplified the stereoscope by replacing the mirrors with prisms cut like a lens . For these instruments, a converging lens with a focal length of about 180 mm was cut into two semicircular pieces, and the two halves, with their circular edges facing each other, were fixed in a frame. Looking behind the lenses, a sheet of paper containing the two drawings (or photographic images) was inserted.
The lens effect made it possible to view the images without the eyes having to adjust to the short image distance ( accommodation ). The prism effect made it possible to use a larger lateral offset than the natural eye relief (about 65 mm) between the two images, which meant that the images could be wider. This in turn made it possible to cover a wider viewing angle and to print or draw the images with higher resolution .
Stereoscopes of this type with a series of paper images were in common use in the 19th century. Usually, however, two small lenses were used, the axes of which roughly coincided with the axes of the eyes (i.e. without prism wedge effect) and pairs of 6 x 6 cm images adapted to the interpupillary distance.
From now on, multitudes of photographers took stereoscopic photos on their excursions around the world. In the British Museum in London, historical stereo recordings of excavations and landscapes on a circular disk are still shown in various rooms. This way of viewing is therefore a forerunner of the popular View Master devices from the 1950s.
In 1851 the French optician Jules Duboscq presented his devices to the public at the World Exhibition in London . They were stereoscopes designed by Brewster, with which he showed stereo daguerreotypes . The response from the audience was overwhelming, and Queen Victoria was also enthusiastic about the presentation. The triumphant advance of stereo images could no longer be stopped.
The stereoscope was most widely used in the design developed by Oliver Wendell Holmes in 1861 , a stereoscope with focus adjustment that became a de facto standard.
August Fuhrmann developed a large circular stereo viewer, the so-called Kaiserpanorama , around 1880 . Around 1900 this became a popular mass medium in Central Europe.
In 1938 Wilhelm Gruber invented the View-Master, a stereo viewer with interchangeable image panels.
Around 1900 as well as in the 1950s, stereo photography experienced a boom. Home stereoscopes have become popular. Publishers offered stereoscopic cards from all over the world. However, due to the higher technical complexity, stereo photography has never established itself in the long term. Today, thanks to the introduction of the digital camera , it is experiencing a slight renaissance because expensive photo paper is no longer needed and experiments are less expensive.
From 1910 onwards, stereo photography was increasingly replaced by the new medium of film.
During the First World War , reconnaissance planes from all warring parties took countless photos. In 1916 they were already operating at altitudes of over 4000 m due to the ever stronger anti-aircraft defenses. With high-resolution cameras and later also series images, they provided important insights from deep inside the enemy. Whole front sections were systematically photographed; The Army High Command created staff picture departments with laboratory, repair and archive facilities. The special series cameras with large focal lengths developed by the companies Zeiss , Görz, Ernemann and Messter were installed vertically suspended in the German machines. Spatially dimensioned image recordings were created using stereoscopic recording techniques, which surveyors and cartographers converted into detailed front maps for the staffs.
When recording a stereography with a real stereo camera with two lenses or a light field camera, you can take pictures as normal. When designing the motif, attention should be paid to a staggered foreground and background arrangement of objects. This promotes the spatial depth effect when looking at the photo later.
Stereo landscape recordings without a foreground rarely appear three-dimensional when recorded with a normal stereo basis (eye relief). Therefore, if you want an excessive space, an extended stereo base is created. For example, a conventional camera is used to take two recordings one after the other, with the camera being shifted horizontally by about 50 centimeters between the recordings, expediently on a slide. A disadvantage of this method is that the object section (the motif) may have changed in the meantime, for example bird flight. This change sometimes disrupts a spatial fusion. It is therefore advisable to take recordings with a wider stereo base with two fixed cameras that are triggered simultaneously using suitable means, for example using a cable release.
A simple recording technique for stereo non-professionals with viewfinder cameras: first object photo with the body weight on the left leg, second object photo with body weight on the right leg. The stereo base affects the deviation between the two photos.
Basic rules of stereoscopic recording
The goal of a good stereo recording is usually a reproduction of what has been seen that is as realistic as possible. Maintaining the same position of the bundles of rays when recording and viewing is the basic condition for a geometrically true-to-life (tautomorphic) reproduction. Otherwise the stereo effect will not be available due to excessive demands or a spatial distortion of the original will result (heteromorphic spatial images).
- The partial image pairs must have the same differences in viewing angles (parallaxes) as for free vision, which is why the recording base should correspond to the mean interpupillary distance of 65 mm.
- When viewing, the same viewing angles as when taking the picture must be maintained. On the one hand, the partial images must be viewed at a distance from the eyes that is the same as the focal length of the recording and must be located in one plane. On the other hand, the distance between the image centers or corresponding distant image points should be 65 mm.
- The image axes of the two partial images must take the same direction when viewing as when taking the picture. This requirement means that for viewing not only, as already required under (2), the partial images have to be mounted at a distance from the taking lenses, but also have to be inserted into the viewer, e.g. a lens stereoscope, so that the lens axes meet the image centers. If the lens centers are shifted laterally with respect to the partial image centers, the spatial impression that comes along appears laterally shifted and distorted, the more the greater the deviation from the normal position, the more so.
- Similarly, if the image and lens centers are not the same in height, distortion also occurs. As long as the height distortion remains within moderate limits and, above all, is the same on both partial images, it is hardly a problem. In contrast, a height difference between the left and right partial image of only a few tenths of a millimeter has the effect of so-called "height parallax" and makes spatial merging more difficult. Therefore, when assembling the stereo images, particular care must be taken to avoid deviations in height of the partial images under all circumstances.
- The differences in position or viewing (parallaxes), which only occur parallel to the connection line of the recording base, must also be parallel to the connection of the lens center when viewed. In other words: The partial images are to be arranged with respect to one another in such a way that their lateral boundaries are aligned parallel to one another and are not canted in their plane to one another. Otherwise, unpleasant height parallaxes arise which have a disruptive effect on the stereoscopic effect.
- The images must be sharp over the entire area of the image recording because the human eye sees all objects at the same time from a distance of about three meters and, on the other hand, immediately focuses (accommodates) at closer distances. "Artistic blurring" is therefore inappropriate in the stereo image and should be avoided. To achieve a good spatial effect, short focal length lenses that have a high depth of field should be used . On the other hand, in contrast to the individual photograph, there is no need to worry about the “correct perspective”. “Real” stereo cameras have a slightly shorter focal length with a large depth of field.
- The space (depth zone) shown in the stereoscopic image should be dimensioned so that it can be captured sharply at once. The difference in viewing angle between the closest and the farthest point must not exceed an amount of 60 to 70 angular minutes - "Lüscher angle". When taking landscape photos, the closest point must therefore only be three meters away. In the case of macro shots, attention must be paid to maintaining the depth zone. Background outside the permitted area should be covered or blurred.
- When viewing the stereo images, the convergence of the visual rays must not exceed the maximum amount of convergence of the axes of the eye by approx.
If you observe the basic rules listed above, you will always achieve a natural and pure spatial effect. Therefore, one need not fear the falling lines that are fearfully avoided in ordinary photography , such as those that occur when taking pictures of buildings with an inclined camera.
Methods of presentation and viewing
Stereo image pair
A simple method is to display two stereoscopic partial images side by side; With a special gaze technique ( parallel gaze - provided that their width does not exceed approx. 65 mm in each case - or cross gaze ) they can then be perceived as a spatial image without any further aids.
No special prerequisites are required to learn how to look at stereo images without tools. To make things easier, however, there are special prismatic glasses. With the KMQ observation method, the partial images are not shown next to each other, but on top of each other.
When archiving stereoscopic image pairs on computers, it is common practice to save both partial images in a single JPEG file and to specify the file type with ".jps". These pairs of images are intended for viewing with the “cross-eyed” (squint).
In the case of the anaglyph images, the two partial images are printed one on top of the other, with both partial images being colored in complementary colors. "Anaglyph" is basically any stereo image in which the two partial images are shown simultaneously on the same surface (the polarization projection is also, strictly speaking, an "anaglyph projection"), but "anaglyphic" usually means a color anaglyphic representation: Zur Separating the two individual images, different color filters are used in 3D glasses , originally red in front of the right eye and green in front of the left. When watching the film, the red filter deletes the red film image and the green image becomes black - the green filter deletes the green color image and the red becomes black. Since both eyes now see different images, a spatial image is created again in the brain.
In the late 1970s, Stephen Gibson significantly improved color anaglyph technology with his patented "Deep Vision" system that uses different filter colors: red in front of the right eye and blue in front of the left. The Danish company "Color Code" now also offers its own color anaglyph system. The filter colors of the “ColorCode” glasses are blue in front of the right eye and yellow in front of the left. Another color anaglyph process (“Trio Scopics”) was introduced in England in 2008 for the feature film “Journey to the Center of the Earth”, with green in front of the left eye and magenta in front of the right.
While red-green and red-blue glasses only use two of the three available color channels of the RGB color space , cyan consists of a mixture of green and blue, which together with the red filter brings all three colors into play (in the case of The same applies to blue-yellow glasses, since yellow is created from red and green light).
A shutter 3D system uses so-called shutter glasses (also known as LCD shutter glasses) when reproducing 3D images. These special glasses have lenses that consist of two liquid crystal surfaces (one for the left and one for the right eye), which can be electronically switched between transparent and non-transparent. This can be used to either darken the left or right eye.
3D polarization system
A 3D polarization system is a method for displaying stereoscopic 3D images. With this method, the images of the stereo image pair are emitted in oppositely polarized light. There are correspondingly offset polarization filters in front of the projection lenses and in the viewer's 3D glasses.
If the 3D image, composed of several individual images, can be viewed spatially on normal photo paper, the help of a special laboratory must be sought. The individual images are exposed in narrow strips on the image carrier and a lenticular film is applied over the overall image , which enables viewing from different angles. The more images are available for this grid film, the less the viewing angle jumps when the image is moved. For this procedure u. a. a special 3D camera developed by the Nimslo company that can even take four photos simultaneously on 35 mm film . Since the 1970s there have been series of postcards (and occasionally large-format pictures) that have used this method.
To view two slides that together result in a 3D photo, two simple slide “peeps” are sufficient for a single person, in which the left and right images can be viewed without further technical effort.
Random point space images
During his research in 1959, Bela Julesz discovered that the perception of spatial depth only takes place in the brain. To do this, he experimented with a special kind of stereo image pairs that only contained randomly distributed points (English "random dot"). The spatial effect is only created by the lateral disparity . A circle can be seen in the following picture.
The principle of random point spatial images is the generation of random point images. The spatial differences are generated as a difference in the second image. The difference in height results from the difference between a point on the first image and its changed position on the second image. This works so well because the brain tries to get the two images to match. It is still completely unclear how the brain recognizes two points on the left and right retina as "belonging together", the so-called " correspondence problem ".
The next development followed with the Single Image Random Dot Stereogram (SIRDS), which is a single large image. This type of stereogram was developed by Christoper Tyler and Maureen Clarke around 1979.
The procedure for creating the SIRDS is similar to that for creating the random point image pair. The difference is that no whole random point image is generated, but a strip first. From this strip a difference strip is calculated, which is attached directly to the original strip; a further difference strip is calculated for the difference strip, and so on, until the entire image is complete. It is advantageous to place the original strip in the middle and to place the difference strips to the left and right of it. This can be particularly understood with the SIS shown below.
In order to get the correct spatial impression, the viewer's gaze must be directed towards the picture into infinity. Squinting gives an inverted 3D impression: image elements that are actually in the foreground appear in the background and vice versa. After a period of getting used to it, the embedded contours become visible.
The computer game Magic Carpet used the SIRDS method on request to show the game in real time. Due to the interactivity of the game, this display is a special case of the animated stereogram.
In the mid-1980s, the random patterns began to be replaced by real images. The single image stereogram (SIS) then experienced in the 1990s, a great boom after Tom Baccei the book series The Magic Eye ( Engl. Magic eye ) brought out.
Stereoscopic motion measurement
Classic stereoscopy records the spatial coordinates (3D position) of corresponding points in an image pair. Many applications require the combination of 3D point clouds to form individual objects. This task can often not be solved on the basis of the 3D information alone. For example, the child who walks into the street in the upper left picture can only be separated from the car in front of him based on his movement. For this purpose, 6D-Vision tracks points with a known distance over two or more successive pairs of images and merges this data. This results in an improved 3D position and at the same time is able to measure the direction and speed of movement for each pixel under consideration. This information (3D position + 3D movement) enables the position of relevant objects to be predicted and potential collision risks to be recognized. The result is shown in the upper right picture. The arrows show the expected position in 0.5 seconds.
The method is also used in the recognition of gestures, i.e. the movement of limbs, without having to model the shape of the person, only using a passive stereo camera.
So-called "Pulfrich glasses" with light / dark filters (e.g. "Nuoptix"), use the "Pulfrich effect" for a 3D impression when tracking shots from the side. B. by the RTL television show Tutti Frutti in the early 1990s. The Pullfrich method is not a real stereoscopic display, since the image is only recorded with a single camera. The two perspectives for the left and right eye come about through the darkened spectacle lens, which is based on the Pulfrich principle. The darkened view is passed on to the brain with a time delay, so that two views from different perspectives (but offset in time) form the spatial impression. This method can only be used to a very limited extent because important requirements must be met for this method to work as a 3D method at all. The camera or the objects must always (fundamentally and continuously) carry out a constant, slow, exclusively horizontal movement. If only one of these conditions is broken, no more 3D effect occurs.
American Paper Optics' ChromaDepth process is based on the fact that colors are refracted to different degrees in a prism . The ChromaDepth glasses contain special viewing foils, which consist of microscopic prisms. As a result, light rays are deflected differently depending on the color. The light rays hit the eye in different places. However, since the brain assumes straight rays of light, the impression arises that the different colors come from different points of view. The brain thus creates the spatial impression (3D effect) from this difference. The main advantage of this technology is that you can view ChromaDepth images without glasses (i.e. two-dimensionally) without any problems - there are no annoying double images. In addition, ChromaDepth images can be rotated as desired without losing the 3D effect. However, the colors can only be selected to a limited extent because they contain the depth information of the image. If you change the color of an object, then its perceived distance also changes. This means that a red object always comes before z. B. green or blue objects.
A number of methods also use the effect that prisms deflect the beam path. For example, B. the stereo viewing device SSG1b, also known under the name KMQ since the 1980s, this effect. Primarily for books and posters where color accuracy and simplicity are important. But it could be used earlier on the screen or for projection with a few viewers. However, the user must maintain the appropriate distance to the picture and keep his head permanently horizontal. Otherwise the lines of sight of both eyes will not coincide with the two partial images which are arranged one below the other. Hence the English name of the process: Over-Under . These restrictions are to be lifted in the future by an open hardware or open source project called openKMQ for working on computers.
In addition to entertainment, stereoscopy is also used to illustrate stereometry and trigonometry , in math textbooks, and to study the laws of binocular vision.
Dove demonstrated the creation of gloss with the help of the stereoscope. If the surface of a drawing is painted blue and the corresponding one of the other is painted yellow, you can see it with a metallic sheen when you look at it in the stereoscope through a purple glass. White and black make the picture even more vivid. Dove also used the stereoscope to distinguish real securities from fake ones. If you look at the papers to be compared with the instrument, you will immediately notice the smallest differences. The individual characters that do not exactly match the original do not coincide and appear to be in different planes.
The horizontal deviation of the corresponding image points on the paired stereoscopic partial images can also be technically evaluated in order to determine the depth. Physiological excessive demands do not play a role here and the effect is used in astronomy , whereby no paired montage of the images is necessary. If, on the other hand, you want a comfortable and natural viewing, perhaps even without major technical aids, then the paired assembly of the stereoscopic partial images to 3D photos is practical and common.
In vehicle and robot technology, stereo video sensors are used to measure distances and distances.
The stereoscopic aerial photo analysis can be used to map terrain formations and to create 3D city models. It was also used in aerotriangulation and photogrammetry until the 1990s .
In technical publications in structural biology , protein crystallography and NMR spectroscopy , stereoscopic images are used to represent three-dimensional molecular structures. These stereo image pairs can be viewed with the parallel view without tools. There are also magnifying glasses for this type of image. It is easy to display molecular structures stereoscopically: A molecule is displayed, rotated by 6 ° in the vertical axis and displayed again. These two images are displayed side by side.
Today, computer games mostly work with three-dimensional models which, with suitable software support, can be displayed not only on a conventional monitor, but also on special stereo 3D monitors with a depth effect. With the help of 3D shutter glasses (synchronized via cable or infrared pulses), the two camera positions calculated in the stereo software are alternately made correctly laterally available to the viewer's eyes, so that a spatial impression of the scene is created in the visual center of the brain. Stereo 3D animation films for digital 3D cinemas are produced in the same way, but with a much higher resolution.
In the case of 3D movie real scenes or 3D television recordings, two high-resolution video cameras are nowadays mounted side by side at eye distance (often only feasible using a "mirror rig") and a "stereographer" (stereoscopy expert) on the 3D image Monitor monitored.
- ↑ http://www.perseus.tufts.edu/cgi-bin/ptext?doc=Perseus%3Atext%3A1999.04.0057%3Aentry%3D%2396561
- ↑ http://www.perseus.tufts.edu/cgi-bin/ptext?doc=Perseus%3Atext%3A1999.04.0057%3Aentry%3D%2395112
- ↑ Double muddled: Fuji Real 3D W1. on: heise.de , July 22, 2009. (Example: heiseFoto introduces a new 3D camera)
- ↑ Hermann Lüscher: The choice of the cheapest basis for stereo distance and close-up shots. In: The stereoscopist. No. 7, 1930.
- ↑ Werner Pietsch: The practice of stereo close-ups. Knapp, Halle (Saale) 1957, DNB 453777589 .
- ↑ Werner Pietsch: Stereophotography. Fotokinoverlag, Halle (Saale) 1959, DNB 453777597 .
- ↑ 3D camera from Nimslo ( Memento from March 7, 2006 in the Internet Archive ) (link in English)
- ↑ springerlink.com: “6D-Vision: Fusion of Stereo and Motion for Robust Environment Perception”, Uwe Franke, Clemens Rabe, Hernán Badino, Stefan Gehrig, Daimler Chrysler AG, DAGM Symposium 2005
- ↑ 6D-Vision.com
- ↑ A. Suppes et al .: Stereo-based video sensors using a stochastic reliability analysis (PDF)
- ↑ Examples in: Duncan E. McRee: Practical Protein Crystallography. Academic Press, San Diego 1993, ISBN 0-12-486050-8 .
- David Brewster: The stereoscope: it's history, theory and construction. London 1856.
- The stereoscope in “Die Gartenlaube” from 1855
- Christian Georg Theodor Ruete: The stereoscope: A popular representation. 2nd Edition. Teubner, Leipzig 1867.
- Fritz G. Waack : stereo photography. 4th expanded edition. Self-published, Berlin 1985, also in English translation due to strong demand from the USA.
- Jean Pütz: The Hobbythek book 3. vgs Verlagsgesellschaft , Cologne 1979, ISBN 3-8025-6102-3 .
- Fritz Waack, Gerhard Kemner: Introduction to technology and handling of 3-D photography. Museum for Transport and Technology, Berlin 1989.
- Alexander Klein, Franz Weiland, Rainer Bode: 3D - but how! From magical images to 3D photography. Bode Verlag, Haltern 1994, ISBN 3-925094-64-4 .
- Thomas Abé: Basic course in 3D images. VfV-Verlag, Gilching 1997, ISBN 3-88955-099-1 .
- Holger Tauer: Stereo 3D. Schiele & Schön, Berlin 2010, ISBN 978-3-7949-0791-5 .
- Leo H. Bräutigam: Stereo photography with a 35mm camera: A practice-oriented introduction to analog and digital 3D photography. Wittig Fachbuchverlag, Hückelhoven 2004, 2nd addendum "Digital Stereo-3D-Photography" 2014, including viewer and data slide, ISBN 978-3-930359-31-8 .
- Leo H. Bräutigam: eBook, 3D photography - 3D video , Civitas Imperii Verlag Esslingen, 2014, ISBN 978-3-939300-28-1
- Achim Bahr: stereoscopy. Rooms, images, spatial images. Thales Verlag, 1991, ISBN 3-88908-549-0 .
- Tom Baccei, Cheri Smith: The Magic Eye. Ars Edition, ISBN 3-7607-2264-4 .
- Marc Grossman: The Magic Eye, Volume I by NE Thing Enterprises. Andrews & Mcmeel, ISBN 0-8362-7006-1 .
- Arthur G. Haisch: Hotel Morbid / Morbid Rooms, stereo room images. 3-D-World Verlag, Basel 1983, ISBN 3-905450-02-X .
- Matthias Henrici, Christian Neubauer: Fantastic moments I. Lingen publishing house
- Hartmut Wettmann: The Rhineland in historical stereo photos. Dr. Gebhardt + Hilden, 1999, ISBN 3-932515-15-3 .
- Ulli Siebenborn: Interactive Pictures, Volume I. Taschen Verlag, 1994, ISBN 3-8228-9211-4 .
- Roland Bartl, Klaus Bartl, Andreas Ernstberger, Peter Schwartzkopff: Pep Art. 3-D pictures of the new kind. Südwest Verlag, Munich 1994, ISBN 3-517-01632-2 .
- Katja Lembke , Arnulf Siebeneicker (eds.); Hannes Wirth: Hildesheim in 3-D. Roemer and Pelizaeus Museum, Hildesheim 2009, ISBN 978-3-938385-29-6 .
- Yuki Inoue, Masahira Oga (Ed.): Stereogram. ISBN 3-7607-1106-5 .
- Andrew A. Kinsman: Random Dot Stereograms. ISBN 0-9630142-1-8 .
- Steinhauser: About the geometric construction of the stereoscopic images. Graz 1870.
- Rolf Sander, Martin Simeth: The little hobbit and the autostereogram In: Spectrum of science . No. 1, 1995, pp. 10-15.
- Imre Pál: Térláttalós ábrázoló mértan. Budapest 1959.
- German: Descriptive geometry in spatial images. from the Hungarian by N. Miklós Marosszéki. German Publishing House of Science, Berlin 1961, DNB 575340851 .