Interface revolution

Source: Venturebeat.com

Heroes of science fiction films rarely use a regular computer keyboard or mouse. It looks much more impressive when a person puts on special gloves and manipulates with holographic images in space, or presses on the touch buttons projected directly on his forearm. However, in recent years, alternative computers control interfaces have ceased to be science fiction and are actively used in game consoles, for example, the Microsoft Kinect system. Modern technologies allow an ordinary user to buy a device that, using a webcam and a microphone, will register its movements in space, and use it to control the computer. But technologies of alternative interfaces continue to develop actively, engineers work on creating a truly accurate or simply radical control method.

The relevance of new interfaces is obvious. For example, recently New Scientist published a report on the first use of the “air” interface in surgery.

Given the limited space of the operating room, the Microsoft Research engineers programmed their Kinect system so that the surgeon could turn over and scaling the X -rays, and the images of computed tomography or three -dimensional models of the organs with one hand and scaling the voice.

Previously, due to the need to maintain sterile conditions, these tasks were performed by assistants, which was not always convenient. In addition, studies have shown that

Using such a system, surgeons can more often turn to the necessary images and “release” attention to complex operations.

Kinect is not a novelty - it appeared on sale at the end of 2010, but only now it begins to fully disclose his potential. However, the resolution of its camera is only 640 by 480 pixels, which are enough to register the movements of the whole body at a distance of up to 5 meters, but too little for high -precision manipulations. The American company Leap Motion said that in the winter of 2012 their LEAP gesture recognition system will be on sale, which records movements within a radius of 20 cm from the sensor, but at the same time has an accuracy of 0.01 millimeter. Unfortunately, the developers do not clarify exactly what technology is used in their device, but if their statements are true, then such a controller can become very useful for 3D models, artists and engineers who are not enough ordinary two-dimensional sensory stylus.

Of course, the case does not limit exclusively to motion recognition sensors. The Disney Research research laboratory, together with the University of Carnegie-Mellon, presented in May 2012 their latest development called Touché, which allows you to turn any object-both a solid object and a human body or even a tank with water-into a sensory cartoon controller. It uses a familiar method for registering a change in the signal frequency, which was previously used in the motion sensors and in wireless communication, however, the Disney Research was the first to guess it to apply it to register touches. Their technology is called Swept Frequency Capacitive Sensing (SFCS), a capacious measurement of the swing of the watch.

A conventional capacitive sensor sends an electric signal to a certain frequency to a conducting object - for example, a finger or stylus - and records a touch on a change in the frequency of the signal. SFCS measures changes at once at several frequencies at once. The human body is heterogeneous and has different resistance in its different parts, and therefore, the electric current passes through it differently, Touché allows not only the fact of touching, but also how exactly the touch of the human body and the registering object were produced. The working prototype can determine the number of fingers with which a person concerns the door handle, the pose in which the person sits at the table, how many fingers a person lowered a capacity of water connected to the system, how many fingers a person put to his own hand and whether the user was an adult or a child.

The system must first be “taught”, that is, to set the frequency ranges for different cases of touch, and high -tech software is already responsible for analysis of the data. At the same time, Touché works with only one electrode connected to the object. Everything that the current can be used can be used, and if the material is non -caring - for example, a tree - it can be covered with a thin layer of metallized paint.

The scope of this technology is unlimited: from improving the accuracy and capabilities of the usual tape screens of telephones and tablets, control of pocket devices using touching your own palm to the complete “electrification” of the home environment-“smart” door handles that include lighting when you come home, or a bathroom that determines it, when an adult or a child was led to it.

According to the developers, they needed a year to create Touché, and documents have already been submitted to the US Patent Bureau.

Another scientist from Carnegie-Mellon, Chris Harrison, together with Microsoft is developing a unique Skinput interface, which uses a human body to enter data. Skinput is based on the principle of sound spread in solid bodies. If you knock on the arm in its different places, then depending on many factors-the thickness and density of bones, connective tissues, muscles-the sound will spread differently. Harrison created a sensor that attaches to the shoulder, and can determine with an accuracy of 95% of the zone in which a touch was produced - five fingertips, a palm, a wrist, various zones of the forearm. At the same time, weather, electromagnetic and other conditions, and even running do not affect the accuracy of the definition. Provided sufficient miniaturization of the detector, it is possible to ensure laboretry control of pocket devices: for example, a double touch of the index finger increases the volume of music in the player, and the double touch of the ring - reduces. And in combination with a PICO projector, you can create interfaces projected onto the hand and controlled by touching hand.

Another Microsoft interface called Omnitouch is also working on a similar principle. It uses a pico projector and a depth sensor fixed on the user's shoulder.

The system allows you to project an interface on any flat surface - a table, a wall, or palm - and register with a depth sensor “pressing” to certain elements of the interface.

This is even more than just recognition of gestures, and to some extent augmented reality. Engineers demonstrated many approximate tasks on the working samples that Omnitouch can perform, for example, can project a “canvas” to the wall, and on the palm of the left hand - a “palette” in which the user selects color and draws on the wall with the tip of the finger; contextually recognize the surfaces that fall into the project, for example, project the main window application on the table, and on the palm - the settings window; Follow the moving surfaces and automatically adjust the projection angle. The system is very bulky, a person has to wear a metal frame on his chest, which is holding a projector and a camera above his shoulder, so there is no need to talk about everyday use. But already now the system can be useful to 3D models or engineers, and with the advent of even more compact projectors and cameras Omnitouch can become mass technology, or at least, find its place in the world of electronic entertainment-as a Kinect system.

The “air” interfaces are unlikely to completely replace the tape screens of phones and computer screens usual in the near future, but definitely the scope of their use will soon cease to be limited to computer games. The smoothness and speed of management they provide can create competition for the usual computer input computer devices - primarily mice - in areas that require naturalness and speed of movement. Probably, the first by the Kinect to a commercial project of this kind will be Leap, and the success (or failure) in the winter of 2012-2013 will show the real interest of the market in such devices.

See also: Microsoft announced the game "Controller without a controller" for Windows