Imprimir Republish


Trick of the eye

Simple equations explain the localization of objects that seem to come from nowhere

EDUARDO CESARThe preferred victims of cursing by soccer fans, mainly when they flag up for offside, when the attacker is face to face with the goalkeeper, linesmen and lineswomen would be thankful if a visual illusion called flash-lag were better known. This phenomenon, which can be looked upon as that which is seen late, manifests itself in situations in which an object in motion is caught by surprise in its trajectory by a second element, which occurs very close to it and abruptly. The scene that comes about gives the impression of the existence of distance, even though it’s small, between the two. What appears by surprise, seems to be behind the first, but in reality the two are in parallel, side by side.

This could an argument to be usesdby the linesmen in order to justify some of their mistakes. On the field, fighting to keep up with the play, in general very fast, the referee’s assistant identifies the presence of the adversary’s last defender only after looking firmly for the position of the attacker and of worrying about the moment of the final forward pass. He remains with the impression that the attacker had been advancing, in front of the defender, and raises his flag, thus marking the offside. But the attacker should have been allowed to play on because the two players had been on the same line – there was no irregularity in the play. “The linesman can easily be deceived by this virtual illusion”, assures the neurophysiologist Marcus Vinícius Chrysóstomo Baldo, a graduate in physics and medicine, and a researcher at the Biomedical Sciences Institute (ICB) of the University of Sao Paulo (USP).

Baldo and Nestor Felipe Caticha Alfonso, from the Physics Institute of USP, based on what had already been known about the working of neurons, created a mathematical model that helps to explain this lateness in perception of objects. The two physicists applied the situations in which the flash-lag appears within classical mathematical approach, already used to study memory and learning, known as artificial neural networks. According this approach, the regions of the nervous system responsible for vision are divided into layers formed by lines of neurons that communicate simultaneously with other neurons of the following layer, but not just in a direct manner: each nerve cells can interact with the others also in a divergent and convergent manner, indeed, making use of diagonal routes. This is the reason for which the information takes more time to flow through this network and, as a consequence, there is a delay in the formation of an image of the object that appears “all of a sudden” – hence the flash-lag. “Up until now the majority of the explanations were conceptual; mathematical models with this biological realism do not exist”, explains Caticha. “We now have an instrument that is more precise and accurate, and that can be tested experimentally.”

The model that they created incorporates physical and physiological explanations, placing in numerical language the connections between the neurons, the electrical stimuli that allow this communication and the very structure of the neural circuits. There are no enigmatic formulae, but essentially only addition and multiplication: the detail is that there are dozens or even hundreds of equations solved at the same time by a computer program. Each equation expresses two mathematical variables: the electrical activity of each neuron and the sum of the influences each one of them receives from the others with which it is related.

The consolidated knowledge about the working of vision was the starting point. We see because the light, on penetrating the eyes, stimulates photosensitive cells, located on the retina, which cover the inside wall of the eyeball and is composed of a complex grouping of neuronal layers. It is on the retina that light and dark transform themselves into nerve pulses, now followed by a second group of neural circuits, the thalamus, a type of nervous system filter that directs the information to a third processing stage, the cerebral cortex. And it is here that the fascinating process of the construction of perceptual vision takes place, which allows us to recognize a player in movement close to the goal, a silent and colored snake in the forest or a familiar face in the middle of a crowd. The physicists’ model incorporates exactly this architecture in layers that characterizes the nervous system, though, in its current stage, there is not the commitment that they identify the model’s layers as being the complex layers of cortical cells or even of the neuronal layers already present in the retina.

As in a theatre
Baldo compares each layer to the seating plan inside a theatre where the places are numbered in accordance with the rows, represented by letters, and the columns identified by numbers. In this manner, a neuron located in the first layer, location A5, interacts with his partners who occupy, for example, the seats B4, B5 and B6 in the second row, which can, on their part, dialogue with their neighbors C3 to C7, in the third row. Furthermore, the electrical impulses exchanged between them – by way of synapses, also simulated by the computer – could be of activation, when the neuron receives an order to do something, or of inhibition, which corresponds to an order not to do anything – as if they were positive or negative signals.

For example, if a neuron were to receive signals coming from five other neurons, three of these signals could be of activation and two of inhibition. If the result of the signals were to be greater than a previously established program value, the receptor neuron would set off a new impulse, which would be transmitted to the following layer. If the sum were to be below this value, known as a limit, a real physiological property, the neuron would remain inactive, without passing on the information. It is this communication based on convergence and divergence, summed to the time taken for its completion, which explains the phenomenon known as flash-lag.

Baldo and Caticha believe that the model applied to flash-lag could help in explaining other visual illusions, such as the Fröhlich Effect, when an object in movement seems to be behind another, static, and hinders the formation of details of the start of the trajectory. Imagine a wildcat jumping out from behind a tree: probably it will not be possible to identify its mouth and snout, which are the first to turn up from behind the tree, and the image of the wildcat would be built up starting from either the left or the right sides of its face. A better understanding about the different visual illusions also represents the possibility of knowing in greater detail the general working of vision itself. Baldo does not discard the hypothesis of any image formed, even those called normal, which in theory do not suffer interferences, could be considered an illusion, since they will never be an exact representation of reality. “I believe that all of us look at the same person or at the same scenario with differences in details, not always subtle”, he suggested. “Vision is always an interpretive reading of the world and there is no absolute precision.”

Late images
Flash-lag began to catch people’s attention in 1958, with an article from the physicist Donald MacKay, from the University of Keele, England, published in Nature. In this work, physicist MacKay described a phenomenon that would remain for many years without an explanation: when he had agitated a lamp and had illuminated it with another source of stroboscopic light – which goes on and off at regular intervals, in successive flashes -, one had the impression of seeing the filament to the front, as if it were outside of the lamp. Only in 1994 did the Indian psychologist Romi Nijhawan, currently at the University of Sussex, also in England, offered the first explanation about the phenomenon, when stating that all of the objects are seen with a delay.

Thus, a car that is coming down a road could already be a meter in front when the brain manages to process the image. According to Nijhawan, the evolution of the human brain must have developed a mechanism to automatically eliminate the difference in phase of space and the delay in the perception of the image, but only when the trajectory is already known. If there is a surprise the brain will not be capable of making these adjustments – and flash-lag will pop up. For this reason the risk of being knocked down by a car that seems to have come “all of a sudden” around the corner is greater.
In 1995, Baldo and the American physicist Stanley Klein, from the University of California, United States, also published in Nature another study concerning flash-lag, showing that this type of visual illusion could well come from lack of attention. The idea was simple: as one’s attention is drawn to the object in movement, it takes a longer time to perceive and determine the position of any new element that appears on the scenario like a flash and calls one’s attention to it.

Neurons in neutral
Very similar arguments were used three years later, in 1998, by two independent groups of researchers: one headed by the American professor of optometry, Harold Bedell, today at the University of Houston, United States; the other one coordinated by two psychologists, David Whitney, currently at the University of Western Ontario, in Canada, and Ikuya Murakami, from the NTT Communication Science Laboratories, in Japan. The two teams had been working with the perspective of time differences in the perception of objects – or latencies. They held that the brain, already accustomed with the previously identified scene, had the need to go through a type of warming up process in order to return to its neural activity and to register a new object. It is as if the neurons were already in a position of resting, in neutral, and thanks to the sudden stimulus they saw themselves obliged to pass again through first, second and third gears until they recover their normal velocity. The model from the two physicists unifies these theories, showing that the proposals previously showing discrepancies or even contradictions are, in truth, facets of the same phenomenon looked at from different angles.

Now, a warning. Before they presented the mathematical model that they created and that will shortly be published in the magazine Vision Research, Baldo and Caticha usually  invite for a test people who visit them for the first time. They ask that the visitor sit in front of a switched on computer and they would off the light in the room, full of archives and papers spread over the desk. With just a click on the mouse a small line appears on the screen, always moving horizontally, in a straight line. When it reaches a fixed point, predetermined and marked, a luminous flash – a second  line – blinks on the screen. The task is to say where this second point had appeared. Most of the times the reply from this reporter was: “Before the other one”. The researchers smiled with satisfaction, faced with yet another victim of the effect of flash-lag: in truth, the two lines had been in line. The explanation that the two physicists then offer clarifies why many times we are betrayed by vision, and allows us to look with more sympathy at the difficulties confronted by linesmen on a football field, under the eyes of thousands of fans.