WiFi signals allow you to take pictures of objects behind walls

He. She

Technological Innovation Website Editorial – 09/15/2023

Take photos using WiFi

Although they have already been extensively explored for detecting the movement of people, the WiFi signals that permeate all of our environments can also be used to image stationary objects, and do so with very good accuracy and resolution.

“Photographing still landscapes using WiFi is very challenging due to the lack of movement,” explains Professor Yasmine Mostofi, from the University of California Santa Barbara. “We then took a completely different approach to solving this difficult problem, focusing on tracking the edges of objects.”

The solution involves the so-called geometric diffraction theory and its corresponding Keller cones, which allow tracking the contours of objects – this diffraction theory, formulated by Joseph Bishop Keller [1923-2016] In 1958, he showed how rays incident on the edge are deflected through space, forming a cone-shaped region.

“When a given wave hits a point on the edge [de um objeto]”A cone of emitted rays appears according to Keller’s geometric theory of diffraction, known as the Keller cone.”

What the team found is that this interaction is not limited to obvious sharp edges, but applies to a wider range of surfaces with very small curvatures.

Killer icons allow you to infer the perimeter of the objects from which their images are created.
[Imagem: Pallaprolu et al. – 10.1109/RadarConf2351548.2023.10149785]

Reading around the edges

The team’s discovery allowed the use of this technology, for the first time, to display and read the English alphabet through walls, using only WiFi signals, a task that until now has been considered very difficult due to the complex details of the letters.

More specifically, the team proposed an image display core based on the Killer icon. This kernel is implicitly a function of edge directions, a relationship that is then used to infer the presence of edges—and their direction, if they exist—by testing hypotheses on a small set of possible edge directions.

In other words, if the calculation determines the presence of an edge, the edge direction that best matches the Keller cone for a given point will be chosen, which will then be used to generate the image.

“The edges of real-world objects have local dependencies,” said team member Anurag Palabrolu. “So, once the high-confidence edge points are found by the proposed image kernel, we propagate their information to the rest of the points using virtual information propagation. This step can further help improve the image, since some edges may be in a blind area, or may “It is dominated by other edges closer to the transmitters.”

Images generated from other letters.
[Imagem: Pallaprolu et al. – 10.1109/RadarConf2351548.2023.10149785]

There is free wifi through the walls

In the demonstration experiments, the team used three common WiFi transmitters. The WiFi receivers are in turn mounted on a remote-controlled vehicle and simulate a network of WiFi receivers as they move. The receiver measures the strength of the received signal, which is then used to create images.

This technology has been extensively tested through numerous experiments in three different fields, including scenarios where images are generated through walls.

The show is entirely valid because it used a challenging scenario, creating images of the alphabet, since the letters represent very complex and diverse details. Finally, they showed how their method allows the use of WiFi to visualize objects behind walls, creating images with enough detail to read letters placed behind these walls. In addition, they also captured images of many other objects, showing that they can capture details that were not previously possible using WiFi.

Other news about: