MSN Virtual Earth to Take on Google

Jim Bruer writes "Microsoft sends news today that founder Bill Gates has announced a MSN Virtual Earth service is to debut in the summer. The service is promised to provide: *Satellite images with 45-degree-angle views of buildings and neighborhoods *Satellite images with street map overlays * Ability to add local data layers, such as showing local businesses or restaurants The service will allow users to choose from a number of different data types plus allow people to contribute their own information."

Comments

Even if Google eventually dies...

[bgarcia]

I hope that Google's legacy is to show people how dynamic web sites should be written. I've never seen a web site anywhere near as responsive as gmail or google maps. Every developer of dynamic websites should study those sites and learn those techniques!

Here are some links to get you started:

• Mapping Google
• Google Suggest Dissected
• Gmail Agent API

I'll assume you know how to find each of the actual google services. ;-)

Actually

[samael]

Actually Microsoft had Terraserver before Google even existed...

MSN had maps longer before Google did

[Carnage4Life]

Considering that Microsoft has had MapPoint and TerraServer long before Google Maps was even a glint in Google's eye, I wonder how exactly MSN is being a copycat.

I guess this is just typical Slashdot pro-Google cheerleading.

Disclaimer: I work at MSN.

Re:either you are a leader or a follower

[Zebra_X]

Last spring there was a demo for Longhorn published to the web. The primary focus was around finding information regarding a real estate property. In the demo the user was able to select the street address of the property and retrieve, photos (air and sat.), demographic information and traffic pattern data and overlay it on a map. In hindsight, this sounds a lot like what Bill is talking about. All of this data was accessible via a web service as well.

I'm thinking that this isn't something that MS just invented "out of the blue" to compete with Google maps, it has likely been under development for a while.

Re:The freedom to innovate!

[hey!]

I dunno, I've been pretty unimpressed with the google map images (beyond such obvious problems as U.S. only coverage) -- the quality is so variable it gets annoying, even for a casual use.

That's what you get with satellite imagery when you order humungous off the shelf data sets.

This problem doesn't take any creativity to fix, just cash. With an outlay of cash you can order up custom imagery that meets your technical specifications with a specified level of consistency. Then you need a staff to check it, and then prepare it for use by rubber sheeting it and registering it, or in this case cutting it up into little tiles. This process requires considerable investment in staff, software, equipment and procedures, but it's efficient once it gets into gear.

Last time I looked into this for a client, I figured he could get really good custom sat imagery for his entire county for something like $10,000. There are 3140 counties in the US, of various sizes (this was a large one); what is more if you're ordering data on this scale, you probably can get a pretty good deal on a per image basis. But we can safely say that if you want really excellent data which fits your purpose precisely and covers the entire US in high resolution, you're talking millions by the time you're done.

Using off the shelf data, you have good enough imagery for a lot less money, which makes sense for the speculative launch of a free service. Now that Google has shown how to use AJAX to make the data more interactive, it's only a matter of time before somebody decides to copy them, but one up them on the data quality. Money seeks obvious problems. Fortunately for Google, they have money too now; maybe they're not fated to being the R&D lab for the industry.

Finally I'd have to say the idea of using images shot from a low angle like 45 degrees instead of overhead is good and bad.

Reasons its bad:

* You can't rectify the image and use it for anything that requires geographic precision.

For example, look at the image in the article, particularly the tower in the upper right hand corner. Consider the column of about 30 windows on the left edge of the tower. The geographic positions of all of these windows are exactly the same, but they show up in different positions in the photograph. The same thing happens when one road crosses another on an overpass. If the angle is such that you can see underneath the overpass, then a geographic position on the bridge deck will have a second representation on the photograph: the point on the roadway directly beneath it. The software which plots the vector representation of the roads is not going to know this, unless the data is tweaked for every overpass in the country. Maybe if you had high res elevation data like a LIDAR survey you could mathematically tweak the entire data set.

People tend to believe a photo more than anything else, but the fact is the precision of photos from a geographic standpoint is highly limited. When using imagery with data from other sources such as GPS and surveying, you can't expect it to line up very well. Things are better if you have in image shot from above with a narrow field of view, and if your target area doesn't have much topography.

* You can't see details that are behind hills or structures.

Obviously. If you are interested in an alleyway that's behind a building, or a lot that is obscured by an elevated highway, then tough.

What is good about the 45 degree image is that it does provide a lot of information that you wouldn't get otherwise about the z dimension, for example you can easily see that in the cluster of buildings on the left side of the image, the building with the pyramid cap is the tallest -- indeed that it has a pyramidal cap. Generally, with imagery, you want one taken in the early morning or late afternoon, especially in low latitudes like Miami. A low sun throws a lot of detail into relief, and a high sun tends to wash it out.