How Google Street View Will Look 5 Years From Now

Google Street View is a feature of Google Maps and Google Earth which provides panoramic views of streets so the viewers can look around at any place that has been covered. What is the difference between Google Street View and handcrafted virtual panoramic photography as of today? What is stopping Street View to look as today’s best VR panoramas? What are the current and future technology constraints? How will Street View look in 5 years?


An example Google Street View image, View Larger Map

Note: The scope of this article is the virtual panoramic photography (and audio/video) part of the Google Street View. I’m not going to tell Google how to do the search thing now.

First of all, let me explain what I mean by virtual panoramic photography. There are many names, some people call it spherical panoramas, VR photography, virtual views, immersive panorama, 360 pano, virtual panoramas, etc. Whatever the name is, the key indicator is that it is interactive, immersive and fully spherical. You can look around as if you were there at the very place.

Hyde Street in San Francisco, Click on the image to open interactive version. Note the difference in quality compared to the Street View image from the same location.

I think of Google Street View (and other companies doing the same thing) as a part of virtual panoramic photography, not it’s competitor. Google Street View has just been one of the key elements that helped pushing panoramic photography to the mainstream during recent years. Handcrafted VR panoramas are usually taken by real people who spend their time shooting the panos and then carefully put them together on their computers. Google has just taken a different approach by streamlining the whole process of taking, processing, storing and putting the images online.

The future

Google Street View car as seen in La Mesa (photo credit goes to kylemurley)

#1 The basics will stay the same

VR panoramic image will probably be created the same way (by Google or by anyone else). In short, you need to take pictures in all directions from your chosen virtual point of view and then put them together by using a special software. That is really all it takes. The number of images you need to take depends on the field of view (or the focal length) of the lens you use. The process of putting the images together is called “stitching”. There’s really nothing more to it. All you need is ANY camera. Yes indeed, there’s a huge difference between just creating a panorama and creating a GREAT panorama. More on this is out of scope of this article so I’ll be posting a complete guide soon on this blog. Make sure you bookmark this article or follow the RSS.

#2 The overall quality will improve

If I could name the most significant difference between a handcrafted virtual panorama and Google Street View, the quality vs. quantity ratio would be it. The very fact that Google Street View imagery is taken from a moving car influences the quality by many ways, some of them are not so obvious. On the other hand, Google can put huge amount of data online this way, far more than could ever be imagined by manual work. I believe that the overall quality will improve in 5 years from now.

#3 Resolution will get higher

Google probably creates higher resolution imagery than it puts online today. And if they are not, they will be in 5 years. The current reasons for leaving the resolution low will slowly dissolve. The time required to process the images will be less of an issue, storage space will no longer be so limiting, end-user computing power for displaying the panoramas will increase. We should be expecting far better zoomability and overall picture quality in 5 years.

#4 Gaps will get smaller

Currently there is a gap between images in Street View. You need to click an arrow to get from one image to another. This gap will get much smaller or it will possibly disappear completely as the image stream will be a continuous flow of images along the street. This will allow video-like driving through cities and looking at objects from any angle.

#5 The car will get more invisible

Many errors and glitches in Street View images today are caused by incorrectly stitched images due to the fact that nobody has invented an invisible car yet. As you can see (or in fact, cannot :-)), the car cannot be seen in Street View images. Google replaces the image of the car at the nadir (bottom) of each panorama with the data taken from surrounding images. In the previous point I said the gaps will get smaller so it will be much easier to do replace the nadir. Or they could just use another sets of cameras and put one in front of the vehicle and another one floating behind.

View Larger Map

#6 Transitions will get smoother

It will be possible to seamlessly go from one picture to another using a technology similar to Adobe’s Infinite Images. Watch the middle third of this video for a short demonstration.

A demonstration of Adobe’s Infinite Images

#7 Glitches will be less visible

The key to understanding quality of Google Street View images is the understanding that you need to take all the images that create the final panorama at the same time and from the same place. This doesn’t sound like a big deal but really the “from the same place” could be a problem. Assuming the cameras aren’t infinitely small they, in fact, cannot be at the same place at the same time. The misplacement of the focal points of the cameras causes a degradation of quality called “parallax error” and it’s usually more visible if there is a contract of distances in objects in the panorama. I believe that parallax errors create a lot of blurry areas on Google Street View images, apart of purposely blurred people faces and car licence plates of course. The only way to minimize it is really is to make the cameras as small as possible. And of course, there is a trade-off between quality and the size of the cameras.

#8 Buildings will get the 3rd dimension

The recent change in Google Street View added a sense of depth. If we take the feature a step further, we can easily end up with 3D buildings and objects. Take these two videos as an example of what is possible today. In five years the 3D buildings are going to be a piece of cake ;-)

C3 Maps

Microsoft Photosynth

#9 It will no longer be street-only

The suitable mass-production pano cameras will get smaller and they will be detached from the cars carried by people. Once again, this is already possible today but the quality is not so great. The new technology will allow to include locations never imagined before in Street View, like towers, cars, boats and even airplanes. Check out my article on 4 Unusual Locations For 360 Panos and 3 Astonishing Aerial Spherical Panoramic Photos to look at some examples of where today’s handcrafted virtual panoramas can be taken. I guess the Street View will have to come up with another name then ;-)

Top of Torre Latinoamericana, Mexico City
(click on the image to videw the interactive version)

#10 Sound will be added

Sound is the “extra added value” some virtual panorama photographers are putting to their pictures today to make the panos more attracting. It allows you to not only see the location you are visiting but also to hear it. Check out this amazing panorama by Jean-Pierre Lavoie for example:

Click on the image to view the interactive version, this panorama has been linked from Jean-Pierre’s site PhotoJPL where he has a great collection of panoramas with sound.

#11 Video will be added

The technology for virtual panoramic video is ready today today however the price/quality ratio hasn’t allowed it to thrive yet.

Pano video example

#12 Live streaming will be added

There will be panoramic video cameras placed at various locations and they will feed live data to the system. We’ll no longer be constrained to seeing only outdated images of the places we visit. The real-word usages are endless, from security and traffic cameras to cameras attached to racing cars, for example.

#13 All people will contribute

The technology will advance so much that the handcrafted virtual panoramic photography will finally merge together with what Google Street View is doing. Don’t get me wrong here. I think there will always be special use cases for pro-level photography. There will always be reasons for doing something with extra precision, differently or in an exceptional quality. However the pano photography will hit the mainstream. There will be point-and-shoot VR photography cameras at a cost of today’s cheapest cameras and the software needed to create panoramas will be widely available, most probably online. People will be able to post their pano pictures and videos, share them easily and they’ll be encouraged merge them with existing images to fill in missing pieces.

#14 A virtual copy of the planet Earth will be created

We will no longer be constrained to view only satellite or aerial photography data and a limited set of panoramas along the streets as today.  The contributions of all people on this planet will create a virtual copy of Earth. You will be able to visit any place and look at it from many different angles. Eventually, when enough data is gathered, it will be possible to choose a month in the year, an hour in the day and apply current or chosen weather to any place on earth you decide to visit. There will be computer games based in this new environment and a lot of businesses will use this data to generate revenue.

How do you see the future of Google Street View and virtual panoramic photography?

If you liked this article please subscribe to RSS. I’ll be posting more articles about virtual panoramic photography soon. Please share this article with your friends, thank you.