Visual Position Navigation, also known as Visual Positioning System (VPS) is the new navigation system developed for Google maps. This visual system uses phone’s camera and Google’s extensive back-end data to analyze surroundings to identify where you are with greater accuracy.
In the last 30 years, visual navigation for mobiles has been a source of countless research contribution since vision navigation can improve the scope of application of independent mobile vehicles. Traditionally vision-based navigation solutions have been devised for Autonomous Ground Vehicles (AGV), but now visual navigation is gaining more and more popularity among researchers for designing Unmanned Aerial Vehicles (UAVs). UAVs offer great perspectives in many applications such as surveillance, patrolling, monitoring, high-risk aerial missions, mapping, fire detection or cinema recordings. Since UAVs move in 3D space they do not have limitations of ground robots which usually cannot overcome rocks, climb stairs or get access to ceilings.
Localization and tracking of a vehicle are major components for providing positions, directions and travel information to its users. These key technologies are very useful for building autonomous navigation systems for mobile platforms. Many approaches have been proposed for locating a mobile robot or a vehicle. Dead reckoning techniques, navigation using active beacons, landmark-based navigation, and global positioning system and vision based positioning are the most commonly used methods.
The main goal of VPS is to show that pure image-based method can be used for mobile robot localization and then investigate its extension to vehicle positioning. This method is different from general landmark based or model-based approaches that depend on known landmarks or prior information about the environment.
A stereo based vision system is built for both extracting features and 3D information of the scene from image sequences. Robust tracking methods based on continuous inter-frame images from the same camera and the stereo image pair from different cameras at a fixed time instant is developed. The relative position of the camera motion is then approximated by registering the 3D feature points from two consecutive image frames.
In today’s world, people find the way using Google maps. But when someone is lost, to find where they are, it becomes a challenge. To solve this problem, Google has launched its own VPS. For example, if a person walks out of the subway, and he does not which direction he needs to go reach his destination, the cameras in his mobile could take pictures of surroundings and then search through its huge databases. When it finds the correct match, Google guides you to your destination. This technology was launched as urban visual positioning and was used in AR City app, which provided a glimpse of AR navigation and maps.
After working on this technology, Google is trying to use this same technology indoors which is called visual positioning using computer vision technology. This technology transforms any indoor environment or space into immersive Accelerated Reader (AR) experience. This technology works by pinpointing a user’s precise location via computer vision and serving up relevant augmented reality content based on that location.
One of the system’s core applications is ‘Indoor Wayfinding’. The technology enables users to navigate indoor environments by providing directions and information directly through their camera in an augmented way. This technology is being applied to large complexes such as airports, shopping malls, stadiums, and supermarkets in a cost-effective manner. These locations can be attached to specific physical locations and presented to users whenever they visit that place.
A good example of this augmented system is, displaying reality menus in front of restaurants in a shopping mall. The menu along with other objects, textures or animations will then appear consistently at the precise location of the restaurant that can be updated in real time.
This combination of these systems can revolutionize many industries like:-
- Shopping & Retail: This will help customers to locate a specific product in huge places like supermarkets.
- Entertainment & Gratification: It can be used to make games like treasure hunting more engaging as they can be placed in exact locations to enhance users experience and enjoyment.
- Tourism: This system could provide an AR tour of places to attract more tourists and could also act as a guide.
- Design: Helps designers in experimenting with new models and style options as they can add AR furniture, art and wallpapers quickly and see the result in their actual space.
This powerful technology can create a revolution. It only needs a camera phone or a standalone camera that could take distinctive pictures. This non-invasive system requires the presence of elements like posters, paintings, aisle which is distinctive and already present in the environment.