{"id":35473,"date":"2019-10-31T22:04:32","date_gmt":"2019-10-31T19:04:32","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novaya-statya-vychislitelnaya-fotografiya\/"},"modified":"2019-10-31T22:04:32","modified_gmt":"2019-10-31T19:04:32","slug":"novaya-statya-vychislitelnaya-fotografiya","status":"publish","type":"post","link":"https:\/\/prohoster.info\/en\/blog\/news\/novaya-statya-vychislitelnaya-fotografiya","title":{"rendered":"New Article: Computational Photography","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

The original article is published on the website Vastrik.ru<\/a><\/noindex> and published on 3DNews with the author's permission. We present the full text of the article, except for the numerous links\u2014it will be useful for those who are seriously interested in the topic and would like to explore the theoretical aspects of computational photography in more depth, but we considered this material excessive for the general audience.  <\/p>\n

Today, no smartphone presentation avoids flaunting its camera. Every month we hear about the latest success in mobile cameras: Google teaches the Pixel to shoot in the dark, Huawei zooms like binoculars, Samsung adds LiDAR, and Apple makes the roundest corners in the world. Innovations are flowing abundantly in this area.<\/p>\n

DSLRs, on the other hand, seem to be stuck in place. Sony showering everyone with new sensors every year, while manufacturers lazily update only the last digit of the version and continue to relax on the sidelines. I have a $3000 DSLR on my desk, but I take my iPhone on trips. Why?<\/p>\n

As the classic said\u2014I went online with this question. They discuss some 'algorithms' and 'neural networks,' having no idea how they specifically affect photography. Journalists loudly announce the number of megapixels, bloggers collectively produce paid unboxings, and aesthetes indulge in 'the sensory perception of the color palette of the sensor.' Everything as usual.<\/p>\n

I had to sit down, spend half my life, and figure everything out myself. In this article, I will share what I learned.<\/p>\n

\u2193<\/a><\/noindex>#<\/a><\/noindex>What is computational photography?<\/h2>\n

Everywhere, including Wikipedia, provides a definition like this: computational photography\u2014any capture and processing techniques where digital computations are used instead of optical transformations. It's fine as far as it goes, except that it explains nothing. Autofocus fits under it, but plenoptic cameras, which have already brought us a lot of useful benefits, do not. The vagueness of official definitions seems to hint that we have no idea what we are talking about.<\/p>\n

The pioneer of computational photography, Stanford professor Marc Levoy (who is now responsible for the camera in Google Pixel) offers another definition\u2014a set of computer vision methods that enhance or extend the capabilities of digital photography, by which a regular photograph is produced that could not technically have been taken with this camera in a traditional way. I adhere to his definition in this article.<\/p>\n

So, smartphones were to blame for everything.<\/p>\n

Smartphones had no choice but to give rise to a new type of photography\u2014computational.<\/p>\n

Their small noisy sensors and tiny low-light lenses, by all laws of physics, should have brought nothing but pain and suffering. And they did, until their developers cleverly found ways to exploit their strengths to overcome their weaknesses\u2014fast electronic shutters, powerful processors, and software.<\/p>\n

\"New<\/p>\n

Most groundbreaking research in computational photography took place between 2005 and 2015, a mere yesterday in the world of science. Right now, before our eyes and in our pockets, a new field of knowledge and technology is developing that has never existed before.<\/p>\n

Computational photography is not just about selfies with neural bokeh. The recent photograph of a black hole would not have come to light without computational photography methods. To capture such a photo with a regular telescope, we would need to make it the size of the Earth. However, by combining data from eight radio telescopes around our planet and writing a few scripts in Python, we obtained the world's first image of the event horizon. It's also suitable for selfies.<\/p>\n

\"New<\/p>\n

\u2193<\/a><\/noindex>#<\/a><\/noindex>Beginning: Digital Processing<\/h2>\n

Imagine if we returned to 2007. Our mom is anarchy, and our photos are noisy 0.6-Mp JPEGs taken on a skateboard. It was around that time we felt the first irresistible urge to sprinkle presets on them to hide the ugliness of mobile sensors. Let\u2019s not deny ourselves this.<\/p>\n

\"New<\/p>\n

\u2193<\/a><\/noindex>#<\/a><\/noindex>Math and Instagram<\/h2>\n

With the advent of Instagram, everyone became obsessed with filters. As someone who once reverse-engineered X-Pro II, Lo-Fi, and Valencia for, of course, research (heh) purposes, I still remember that they consisted of three components:<\/p>\n