Virtual reality technologies seem to be something new and not very common today. However, the first attempts to make such systems took place in the sixties of the last century. Then the first generation systems were more like a modern computer, rather than what we currently refer to as VR. The man there was only an observer and he was not allowed to interfere in another reality. Compared with the first steps, modern solutions are actually not so far away and advanced. Although today the user can fully immerse visually into VR, move within and interact with it, the system still is more like the first generation one rather than the system that we want to see, namely, the one completely replacing our reality. The available solutions only improve characteristics without introducing drastically new functions into the system. There are a number of their shortcomings: limited boundaries, the need for a wired connection to a computer, the impossibility of sharing the same VR space with another user, limited interactions with the different reality and, above all, the impossibility to fully sense a virtual reality with touch.
Today the market is represented by dozens of solutions in the field of virtual reality technology, but they are all similar to each other, differing only in minor ways. The general structure of such solutions can be described as follows: goggles displaying a virtual world measure technical information about the location of the user. Goggles are connected to a computer where all of the necessary computations for visualization are rendered. So-called “controllers” are used for interaction with a virtual world. They are special devices which also measure their position in space and transmit this position data to a computer – this is how interaction with virtual reality is done. However, such a system in itself is not very accurate and its errors will be visible to the user, therefore additional external devices are used to correct all measurements. These devices are basically a system of several laser rangefinders by which the location of the user and controllers is more precisely determined. As can be seen from the description below, such systems are characterized by the following drawbacks:
1. The goggles are connected to a computer most often by a wire, which significantly limits the freedom of user’s movement.
2. The interaction with virtual reality items is limited by the use of controllers. This does not allow one to feel a different world to the full extent.
3. Expensive laser rangefinders greatly increase the cost of the whole system as well. These devices require precise adjustment and calibration, which causes inconvenience to users. In addition, this approach severely limits the range of the systems.
4. The fundamental impossibility of working with multiple users simultaneously.
Rather than simply concentrating on improving system’s performance we focused on expanding the functionality and solving the existing problems while developing our system «Uniblock». Our main task was to immerse the person most deeply into the other world, giving him the opportunity to become an active participant in it. As a result, we want to free the person in movement, let him feel and change this new world. To achieve these goals we had to look for non-standard solutions. Our system is relieved of the need to use expensive laser technology, instead of which a simple camera and modern algorithms are used. Just by applying modern solutions in the field of computer vision, we got rid of hardware controllers, replacing them with full control via hand gestures. Transfer of a human into virtual reality world is done via a sensor system which forms a human skeleton. Due to this approach we opened a possibility of a multi-user mode in our solution.
DESCRIPTION OF SYSTEM MODULES
When developing the system we immediately abandoned the use of wires at the stage of modeling. We managed to transmit without wires not only technical data but also video data, which allows to completely disengage the person from a computer ensuring his freedom of movement. Also due to the rejection of wires in the sensor system, it was possible to obtain such a flexible system that it allows transferring to the virtual reality not only the person, but also absolutely any real object. This fact increases the immersion of a person in the virtual reality world due to the possibility to contact with real objects around him.
In our project we abandoned positioning through the laser rangefinder system and used a stereo pair (a system of two cameras). Due to this approach and several algorithmic solutions we were able to solve the problem of positioning objects and users at a lesser cost. Although the entire workload falls оn a computing device we have focused on the fact that computers are getting cheaper every year and in laser technology this trend is not observed. We also consider the use of specialized computing devices to remove the workload from the user’s computer.
Guided by the same principles we have established a system of sensors, which allows not only to bring humans and objects from our reality into the virtual one, but also has the potential to move virtual objects into our reality. This happens due to the fact that the data analysis of the sensor position occurs for each sensor independently, and only then the data is combined into a «skeleton». Such a solution allows us not to confine ourselves only to the «skeleton» of a person but to transfer absolutely any design to a virtual reality. Also, the use of self-organizing network technology allowed the addition of new sensors to the already operating system with maximum ease.
These technologies may be applied not only in virtual reality but also in augmented reality systems, which opens prospects for their use in this area. Also, the chosen system architecture allows its use not only in the closed areas where most of the known analogues are now working, but also in the open space area.
Work on our product continues on a daily basis and we will achieve even greater opportunities for our future users.