80% of human external sensory information comes from the visual system. Our daily reading, the recognition of the spatial distance of surrounding objects, etc., are closely related to the visual system. However, we are also well aware that “seeing is not necessarily true”, because this visual-brain system is too easy to be deceived. Such as common optical illusions. The most interesting point is that even if you later learn the truth about optical illusions, the brain will still make the same mistakes…In other words, we can’t overturn the stubborn illusions built in the brain.
A man dressed as a miner jumped onto the stage with a light overhead and a pickaxe. After swinging the pickaxe back and forth several times, he claimed to have discovered a “super magnet”-its magnetic force is so strong that it can even attract wood. The screen above the man’s head seemed to confirm his statement: a few wooden balls were rolling up along four slopes, as if free from the restraint of gravity.
Surprise ball: The wooden ball can roll up along the slope. This is an optical illusion work designed by Kokichi Sugihara, which has won awards.
This “miner” is actually Sugihara Atsushi, a mathematician at Meiji University in Tokyo, Japan. He was participating in the finals of the 2010 Best Optical Illusion Competition held in Naples, Florida. As the video continued to play, people found that the balls did not really roll up. Viewed from the front, the pillars supporting the ramp are vertical, but when looking around to the back, it is not. In addition, the ramp that seems to slope downward from the intersection point is actually tilted.
This design using optical illusions (won the highest award of the year) was not done by Sugihara alone. He also had an unusual partner: a computer program.
In the beginning, Sugihara didn’t think about becoming a master of optical illusion. In 1980, this young mathematician was only interested in robot vision and computer-aided design. For this reason, he wrote a polygonal line drawing to generate multiple three-dimensional graphics (when these three-dimensional graphics are projected on a plane, this A computer program that will appear as a polygon)-This is a bit like an object that can be projected from a shadow backwards.
To test the program, Sugihara imported portraits of objects that could not exist in the three-dimensional world, such as the endless stairs of MC Escher . Unexpectedly, in some cases, the program did generate very similar three-dimensional objects based on this. He said: “I thought there was a bug in my software.”
After careful observation, he realized that the assumption that “impossible figures” do not exist in the three-dimensional world was wrong. So Sugihara began to use programming to design paper models. Gradually, he found that his program was not suitable for robotic vision and computer-aided design research. He began to devote himself to exploring all kinds of strange structures, these confusing structures are easy to make with programs. He built an illusion maker .
In the following 34 years, Sugihara designed more than 100 works with the illusion maker: for example , physical models of various impossible figures (such as the stairs drawn by Escher) , and counterintuitive movement patterns (such as a small ball rolling up ) . Arthur Shapiro , a neuroscientist at the American University of Washington , said: “Sugahara systematically created the world after the collapse of the visual system.” Shapiro is both a scientist and an artist, and he aims to reveal how the brain builds the world. The underlying mathematical knowledge.
Regarding the visual information received by the eyes, the human brain usually selectively ignores many possible interpretations. The brain has limited energy and needs to quickly understand visual information, so it cannot accept all weird interpretations. It will only rely on past experience and internal visual processing mechanisms to find the most likely explanation for the rationalized information.
Susana Martinez-Conde, a neuroscientist at the State University of New York State Southern Medical Center in Brooklyn , hosted the illusion contest with her colleague Stephen Macknik . She said that in most cases (though not always) , out of practical needs, the brain’s interpretation is close enough to the real world. “From an evolutionary perspective, if we want to understand information 100% correctly, the cost to humans is much greater than it is now.”
Sugihara believes that the reason why his optical illusion works are confusing is simple: even if a right angle does not exist, people will automatically perceive a right angle. In many of the most compelling impossible figures, some structures consist of only three lines oriented in different directions. In this case, people seem to be tempted to think that these lines are perpendicular to each other.
Sugihara verified this hypothesis by instructing the program to select the three-dimensional graphics with the most right angles. He said that the computer usually selects the three-dimensional figure that people perceive. When there is more than one three-dimensional figure with the most right angles, Sugihara guesses that the human visual system will decide how to perceive it according to the changes in light and shade. He said that the visual system tends to interpret lighter surfaces as upwards and darker surfaces as downwards. He hopes to build a module in the test program to verify this conjecture.
McNick said that once we see it as a right angle, even if this perception has no meaning, we tend to stick to it. We can only observe and process every detail of the image from a local perspective, which leads us to “see” structures that do not exist. He said: “The reason why we think that Escher’s ladder is always upward is because from a local perspective, the angle between adjacent ladders is too reasonable. We cannot observe the whole picture, and therefore cannot detect the small deviations that lead us astray, which is why we can see impossible figures. “
Part of Ascending and Descending, MC Escher, March 1960. © Dreamstime
Sugihara’s ability to systematically create optical illusions provides new ideas for exploring the visual system. He said that some psychologists admire another visual theory: that is, the visual system pursues symmetry when processing information.
Sugihara has started to work with Akiyoshi Kitaoka , a psychologist who studies illusions at Ritsumeikan University in Kyoto, Japan , to figure out what kind of optical illusion designs his can effectively distinguish the theory from the right angle theory. The results of their work will help researchers prioritize various visual shortcuts when we look at things. Martinez-Kant said that the lower the visual level of illusion processing, the more difficult it is to eliminate the illusion. The depth illusion (which is reflected in Sugihara’s works) in the middle-level visual processing is “very stubborn.”
Sugihara’s high productivity is not only the result of computer programs, impossible graphics are also the source of inspiration for his creation. He often moves a part of an object from the background to the foreground to create objects that do not exist in the real world. In addition, he will first cut and then reorganize the graphics, so that all angles still look vertical, but the new image does not exist in the three-dimensional world. Using these simple methods, Sugihara created all kinds of impossible figures. Most of them cannot be turned into three-dimensional objects, but his program can distinguish some objects that can exist in the real world.
Cabin: The top two figures were cut in half and recombined to create two impossible figures. © Sugihara Atsushi
In recent years, Sugihara has begun to show impossible motion patterns-three-dimensional figures are real, but the objects moving on it seem to violate the laws of physics. Sugihara’s program can create three-dimensional figures with straight lines and planes. Analyzing the movement that occurs on the three-dimensional graphics is simple, and Sugihara has designed such a module. For example, in order to create the optical illusion of a small ball rolling up a slope, Sugihara simply drew a sketch of the ramp device, and then asked the program to find all the mappings in the three-dimensional world, and then identified three-dimensional objects that could make the ball “roll up”. Graphics.
When we see these impossible objects or movements, our cognitive consciousness is not strong enough to overturn the brain’s reflective interpretation of three-dimensional geometry. Sugihara said: “The interpretation of three-dimensional images cannot be controlled by rational logic.”
Folding ladder: One of the optical illusions designed by Sugihara-a pole seems to be passing through the folding ladder, moving in a seemingly impossible way.
The impossible movement is evident in Sugihara’s latest work, and he has agreed to show it in this article. As shown in the video, a straight stick is passing through a structure similar to a folding ladder, but in fact, if the stick cannot bend, it is impossible to move in this way.
In order to create this optical illusion, Sugihara first drew a sketch of the ladder, and then used the program to generate and select an unexpected three-dimensional figure. Our preference for right angles makes us think that the top of the ladder is flat, so that the top of the ladder is perpendicular to the supporting rung. But in fact the top is not flat: a few rungs are slanted, so the small stick can pass through the folding ladder in an unbelievable way.
Sugihara’s program may make people think that there are sophisticated computer vision systems in the world that can not be deceived like humans. However, Dale Purves , a neuroscientist at Duke University in Durham, North Carolina , said that without evolution, any computer program is unlikely to observe objects like humans. He said: “Our vision system has benefited a lot from millions or billions of years of evolution. What you see depends on how you want to see it. I think unless through evolution, it depends on engineers. According to the logic, machines cannot observe objects like humans.”
If robots can really penetrate the essence of evolution, they may inevitably produce illusions like humans. The neuroscientist Beau Lotto of University College London and David Corney of Robert Gordon University in Aberdeen, Scotland invented through trial and error to learn how to determine the gray value of synthetic natural scenes. Robot.
Like humans, these robots are also deceived by the “white illusion”. Bright shadows have exactly the same gray scale, but those close to the black stripes look darker, and those close to the white stripes look lighter. However, the computer vision system that has been honed by evolution may also be combined with Sugihara’s program to achieve the best of both worlds. McNick said: “I am sure that computer vision systems will eventually be better than human vision systems.”
At present, the programs of Sugihara and others are not as good as the human visual system. But they are not affected by the preferences of the human visual system, which is exactly what Suigahara intends to study to the greatest extent possible. He said: “I am very happy to use optical illusion works to surprise people.” These works are not only derived from professional interests.
Posted by:CoinYuppie，Reprinted with attribution to:https://coinyuppie.com/the-master-of-creating-weird-optical-illusions/
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.