The first half of photogrammetry

Photogrammetry is a subject that has been developed for more than one hundred years. Its origin predates computer vision. Literally, photogrammetry can be divided into "photography+measurement", so its development can also be divided into the origin and development of photography and the measurement technology developed on the basis of photography. Judging from its English name, the name photogrammetry originated from the Greek words "photos" (light), "gramma" (something written or drawn) and "metron" (measurement). The combination of these words reveals the development path of the subject invisibly, and the existing perception of "photo" (light) phenomenon is further applied to painting.

It all started with pinhole imaging. It is said that more than 2000 years ago, it was found that blocking the wall with wooden boards with pinholes would form the reflection of the object on the wall. If the middle board moves back and forth, the size of the image on the wall will change accordingly. This small discovery became the kindling of the development of photography history.

After discovering pinhole imaging, smart people designed a camera called black box (Figure? 1), you can project an image on the screen. Specifically, the black box is actually a sealed box with a small hole on one side, and the scenery outside the box passes through the small hole to form a reflection on the completely dark inner wall of the box. Smart readers can find that this is actually the earliest form of camera.

Because of the need to project images, in order to make the images clear, the environment must be relatively dark, so many black-box experiments in history were carried out in dark rooms. Later, the black box itself was improved step by step and became a smaller and more portable box, which became a good helper for painting assistance widely circulated in the painter circle.

Due to the limitation of black box (after all, it is inconvenient to carry a big box every time ...), some artists have thought of different painting methods, which can make them present a work better, such as Alberti. 1435, Alberti went to Florence with Pope Eugene IV. Through contact with local artists, he created a theoretical work on painting as an artist and completed it in Latin. In this book, Albert put forward a new method of constructing three-dimensional space on two-dimensional plane by monocular perspective-perspective method. In fact, this has developed from the application of simple natural phenomena to a scientific theory (think about the scientific development in China, what is missing is the step from wonderful skills to scientific theory? 0.0)。 Subsequently, krupa drew the topographic map of Swiss mountains with perspective method, and Bo Temps-Beaupre drew the map of Santa Cruz Islands by measuring angles and distances on the basis of perspective method (figure? 2)。

As we all know, the great Leonardo da Vinci summed up the experience of his predecessors and tried to determine the painter's line of sight from perspective painting (Figure? 3), combined with the black box, 1500, Leonardo da Vinci returned to Florence and created the Mona Lisa by various painting methods such as perspective.

With the development of theory and the expansion of application, many researchers have thought of adding various lenses to the black box. 1540, gerolamo cardano put forward the idea of adding lenses. In A.D. 1568, the Italian aristocrat Daniel Barba Rowe reformed the previous lensless camera box, installed a convex lens, and became the first inventor of the lens camera box in history. He also recorded the principle of the lens camera box by the near-far practical method: tie a rope to the small hole of the camera box, and the small hole can be enlarged or reduced at will, and a very clear image can be obtained after adjustment.

Later, Dante, an Italian mathematician and astronomer, published the idea that the inverted image can be turned into a positive image in Euclid's Law of Distance in A.D. 1573, which greatly improved the photography method (finally, the inverted image is no longer used! ! )。 By 16 1 1 year, Kepler used the composite light transmission of concave lens and convex lens in Diaptria for the first time, which made the image in the black box show unprecedented photographic resolution, and the world respected him as the ancestor of photographic optics. In the same year, lucida, the brightest star of the camera, came out (Figure? 4) People can get better works through the brightest stars.

With the birth of the camera, people can finally use these technologies to shoot works. For example, in 1825, Nipps, a Frenchman, entrusted Charles Sefer, a French optical instrument manufacturer, to make an optical lens for his photographic cassette, and the following year he put the photosensitive material he invented into the cassette to shoot and record the first photographic work in history (Figure? 5)。 His work was successfully shot in the window of his attic in Burgundy, France, and the exposure time was more than 8 hours.

1839, silver disk photography came out, silver disk photography (figure? 6) It was invented by Daguerre, the chief landscape painter of a famous opera house in Paris, and the exposed silver salt coating was developed by mercury vapor. The exposure time of this photography is about 30 minutes, which is much shorter than that of Nieps solar hardening photography. And the photos taken by this method have the characteristics of delicate shadow lines, uniform tone, difficult fading, unable to copy and opposite images. This method of photography is named after the silver plate photography, so it is also called the silver plate photography.

By this time, the technology of photography has made great progress, and people began to think about whether it can be measured by photos.

185 1 year, A.Laussedat of France first created photographic instruments and working methods suitable for photogrammetry, and drew a map of Wansen Castle, marking the birth of photogrammetry. 188 1 year, based on two photos of buildings, Maidenbauer conducted the first photogrammetry experiment of buildings by intersection method, and in 1885, a Prussian photogrammetry photo library was established as the building image archives in Berlin. 1889, the first photogrammetry manual produced by popular science was published, which made more people come into contact with photogrammetry and let more people begin to understand photogrammetry. 190 1 year, the German C.Pulfrich created the stereo coordinate measuring instrument, and the ground photogrammetry developed into the ground stereo photogrammetry, that is, the technology of using the stereo images taken from the ground to build a model on the ground stereo mapper and then mapping the terrain. It is suitable for surveying and mapping in mountainous areas, small mountainous areas and hilly areas, and can also be used for exploration in geology, metallurgy, mines, water conservancy and railways, and the measurement technology is further improved.

The progress of ground photography makes people intuitively think of aerial photography. At this time, someone thought of putting a miniature camera on the pigeon to take pictures in the air (Figure? Seven pictures? 8) The photos taken in this way can be significantly richer, wider and easier to study!

French photographer Nadal took the world's first aerial photo on a hot air balloon more than 80 meters outside Paris in 1858. Subsequently, Blake took an aerial photograph of Boston with a wet board at 1860. Twenty years later, on 1885, the French took aerial photos of Paris on a 2000-foot balloon. Then 1906 San Francisco earthquake caused a fire, and Lawrence took photos from the air with a 17 kite hung by a giant camera. Wei Siddall Shamflug invented Shamflug's law (Scheimpflug principle) in order to use his perspective plotter in aerial photography. At the beginning of the 20th century, the Wright brothers invented the airplane, and the airplane manufacturing industry began to stand out, making aerial photogrammetry possible. 19 1 1 year, Wei Sidar Sham Frog realized the first full-page photo of aerial photogrammetry. The development of aerial photogrammetry has brought photogrammetry to a new height.

Of course, it must be said that mathematics has been running through photogrammetry, and its appearance has solved many problems, such as 3D composition, epipolar geometry, trifocal geometry, motion estimation, 3D modeling of a single image, uncalibration, 3D reconstruction of an image, etc. Projection reconstruction and camera self-calibration also promote the progress of photogrammetry. The development of these technologies will be described in detail in the second half of our article. Let's follow them closely ~