Google will use global localization to improve Maps navigation
Google Maps is one of the most used navigation tools in the world. Yet, it's not always perfect. Have you ever wondered which direction you need to head in or went the wrong way first? Google admits that improving the accuracy of the blue dot has been a priority for them over the years. Now the company is experimenting with a technique called global localization.
It will enable you to determine where you need to go much more quickly. The technique utilizes VPS (Visual Positioning Service), Street View and machine learning to more accurately identify position and orientation. Previously, Google mostly relied on tools like GPS and compass, however it found that "both have physical limitations that make solving this challenge difficult, especially in urban environments".
GPS, for example, relies on measuring delay of radio signals from multiple satellites to determine a precise position. In big cities and dense urban areas that becomes a problem due to low visibility to the sky and signals reflecting off of buildings. This, in turn, results in an inaccurate position on the map.
VPS, on the other hand, determines location based on imagery. It "first creates a map by taking a series of images which have a known location and analyzing them for key visual features, such as the outline of buildings or bridges, to create a large scale and fast searchable index of those visual features." Of course, this means that its accuracy is affected by the quality of the imagery. This is where Street View comes in.
It is an extensive source of high-quality photos of global locations. VPS connected to Street View data can make use of information gathered from across the globe. There still are challenges, however. Street View photos can be outdated - urban environments can change very quickly.
This issue is being addressed by machine learning. Google uses it to filter out temporary parts of a scene and focus on permanent structures that are unlikely to change over time. The company also claims that this is just one of many ways they use AI to improve accuracy in Maps.
Global localization is also being used in the new AR navigation feature, which begun rolling out to select Local Guides. Google will continue developing the technology and sees great potential "for smartphone cameras to add new types of valuable experiences."
What do you think? Are you excited about the improvement of urban navigation? Let us know in the comments.
If they're using street view it won't work in all areas. Where I live, most areas have no street view available and where it is available it's often years old. Hopefully this means they'll be investing in more updates to the underlying map data not just the AI and app functionality.