22 April 2005

Google satellite maps combined with ground based images to form user-friendly panoramic maps


Quick tips

  • A9 Panorama: Ground-level!
  • 3D Google maps: Beta. Wow!
  • Save your Google Maps
  • Technorrati links for Satellite images of landmarks
  • Using Firefox to create a multimedia effect in Google Maps
  • Google Maps Hacks Wiki


  • In Feb 2005, I blogged about some XML concepts related to integrating still images into 3D panoramic maps: XML Reflection; XML Foto-location; XML Photo Analysis; and XML Image Convert.

    Looks like things are getting closer to that vision. It’s nice to see a map come alive with actual images.

    Google satellite image landmarks. Google maps offers satellite images; as does Globe Explorer. Also, zoom in on Swiss cities and streets with MapSearch.

    Obviously, a satellite map is from above.

    But other than people who are in the air, how many people do you know who are using a map on the ground want to see what the top of a building looks like; and does a satellite image from above really help you know what to look for while traveling?

    There are two different points of view. When I walk or take the train, I’m not in a satellite looking down. I’m on the ground looking horizontally.

    So that’s what I’d like to see in the directions and map tools. And I thought there might be a day when we see a different angle in these images: Not just from above, but horizontally. And using images that others had taken.

    The current satellite image brings the location alive with full color. Going further, the next step would be to integrate the map and travel directions with a tool that identified the ground-level-images of landmarks.

    Search engines can index images based on the name. Why not go one step further and apply this image-indexing mechanism and integrate it with the maps, directions, and satellite images.

    In other words, it would be far more interesting to pick a location in a search engine, get a map, and then find the actual image of the places and landmarks I would want to look for while traveling along the route.

    No longer would we have to look for signs or remember how many miles we traveled; rather, we could simply remember to turn left at the big building.

    Going further, what if the map-image was not simply a satellite image from above, but an integrated picture formed from publicly available ground level images?

    What could be possible

    I think a far more useful approach would be to integrate Google Maps with Flicker like this:

  • Users could click on their location or destination or area of interest

  • Users would be given a top view from a satellite

  • Then users would have the ground-view of the buildings.

  • Users would then be able to see a panoramic view of that location based on an integrating tool that combined the available images into a single image.

    This is where the users would look at the surrounding locations not form above, but from the angle that they’re most likely to be when they are on the ground, walking.

    Ideally, the images could be based on geo-location. So that when someone searches in Yahoo or Google maps, the user would be able to see the satellite map first, then zoom into the ground-level view from Flickr.

    Images chosen based on time of day

    During the day, the lighting changes. Sometimes a place that looks one way with one set of shadows, can take on a whole new look as the day progresses.


    Two examples


  • Sextant

    The image-hunting-tool would determine which direction the image was taken based on the lighting, angle of the sun using a reverse-sextant [that thing the mariners use to measure the distance from the horizon to a star]. The image-analyzers would track the angles of the sun in the image, and then back-calculate the time of day.

    Then the tool would be able to select from those candidate-images those that were closest to the time of day the user is traveling. Now you’ll know what to really look for when you drive near that big building that appears to change colors as they day passes.

  • Moon over Desert

    There’s another analogy for converting an image to time of year. They did the same with Ansel Adam’s image of the moon over the desert.

    Based on the angle of the moon and the lighting they were able to determine the point the image was taken, and also the time of year and exact time the image was taken.


  • The next step would be then to connect the many versions of a similar location, and create a combined panoramic view of the area of interest.

    Imagine being able to state the actual time you were going to travel and have Flicker integrate with Google in providing the types of lighting you would see at the time you were traveling. No longer would you get confused about a location because the time of day was different than what you saw in the image.

    Create a composite panoramic picture

    Then the next step would be to develop a mechanism that would integrate these many images in to a composite panoramic picture of that location.

    A9 has a nice version of this panorama-ground-based approach here.

    Think of this as an online version of Photoshop: The tool would take all the images related to an area of interest, and create a composite-panoramic picture of that location or building using all the available images.

    Ideally, this panoramic picture would be a sum greater than its parts and provide an image of a location that was fully integrated with the geo-location mechanism that are associated with blogs; and also pick and choose like the search-engine-image-selector in choosing the best images to make the final panoramic view.

    Summary

    In the end, what would be nice to see is something that not only showed the satellite view, but was a tool that showed me the ground level view of that location.

    There may be other solutions to do this. But an “easy” approach would be to develop a system that indexed Flickr Photographs, and then let users pick a location, and have these ground-level images form a panoramic view of the area of interest.


    LEGAL NOTICE


    Creative Commons License

    This work is licensed under a Creative Commons License.

    You may not copy any of this work to promote a commercial product on any site or medium in the universe.

    If you see this work posted on a commercial site, it violates the creative commons license; and the author does not endorse the commercial product.

    Free to use for non-commercial uses. Link to this original blogspot and cite as .

  • Quick tips

  • A9 Panorama: Ground-level!
  • 3D Google maps: Beta. Wow!
  • Save your Google Maps
  • Technorrati links for Satellite images of landmarks
  • Using Firefox to create a multimedia effect in Google Maps
  • Google Maps Hacks Wiki


  • In Feb 2005, I blogged about some XML concepts related to integrating still images into 3D panoramic maps: XML Reflection; XML Foto-location; XML Photo Analysis; and XML Image Convert.

    Looks like things are getting closer to that vision. It’s nice to see a map come alive with actual images.

    Google satellite image landmarks. Google maps offers satellite images; as does Globe Explorer. Also, zoom in on Swiss cities and streets with MapSearch.

    Obviously, a satellite map is from above.

    But other than people who are in the air, how many people do you know who are using a map on the ground want to see what the top of a building looks like; and does a satellite image from above really help you know what to look for while traveling?

    There are two different points of view. When I walk or take the train, I’m not in a satellite looking down. I’m on the ground looking horizontally.

    So that’s what I’d like to see in the directions and map tools. And I thought there might be a day when we see a different angle in these images: Not just from above, but horizontally. And using images that others had taken.

    The current satellite image brings the location alive with full color. Going further, the next step would be to integrate the map and travel directions with a tool that identified the ground-level-images of landmarks.

    Search engines can index images based on the name. Why not go one step further and apply this image-indexing mechanism and integrate it with the maps, directions, and satellite images.

    In other words, it would be far more interesting to pick a location in a search engine, get a map, and then find the actual image of the places and landmarks I would want to look for while traveling along the route.

    No longer would we have to look for signs or remember how many miles we traveled; rather, we could simply remember to turn left at the big building.

    Going further, what if the map-image was not simply a satellite image from above, but an integrated picture formed from publicly available ground level images?

    What could be possible

    I think a far more useful approach would be to integrate Google Maps with Flicker like this:

  • Users could click on their location or destination or area of interest

  • Users would be given a top view from a satellite

  • Then users would have the ground-view of the buildings.

  • Users would then be able to see a panoramic view of that location based on an integrating tool that combined the available images into a single image.

    This is where the users would look at the surrounding locations not form above, but from the angle that they’re most likely to be when they are on the ground, walking.

    Ideally, the images could be based on geo-location. So that when someone searches in Yahoo or Google maps, the user would be able to see the satellite map first, then zoom into the ground-level view from Flickr.

    Images chosen based on time of day

    During the day, the lighting changes. Sometimes a place that looks one way with one set of shadows, can take on a whole new look as the day progresses.


    Two examples


  • Sextant

    The image-hunting-tool would determine which direction the image was taken based on the lighting, angle of the sun using a reverse-sextant [that thing the mariners use to measure the distance from the horizon to a star]. The image-analyzers would track the angles of the sun in the image, and then back-calculate the time of day.

    Then the tool would be able to select from those candidate-images those that were closest to the time of day the user is traveling. Now you’ll know what to really look for when you drive near that big building that appears to change colors as they day passes.

  • Moon over Desert

    There’s another analogy for converting an image to time of year. They did the same with Ansel Adam’s image of the moon over the desert.

    Based on the angle of the moon and the lighting they were able to determine the point the image was taken, and also the time of year and exact time the image was taken.


  • The next step would be then to connect the many versions of a similar location, and create a combined panoramic view of the area of interest.

    Imagine being able to state the actual time you were going to travel and have Flicker integrate with Google in providing the types of lighting you would see at the time you were traveling. No longer would you get confused about a location because the time of day was different than what you saw in the image.

    Create a composite panoramic picture

    Then the next step would be to develop a mechanism that would integrate these many images in to a composite panoramic picture of that location.

    A9 has a nice version of this panorama-ground-based approach here.

    Think of this as an online version of Photoshop: The tool would take all the images related to an area of interest, and create a composite-panoramic picture of that location or building using all the available images.

    Ideally, this panoramic picture would be a sum greater than its parts and provide an image of a location that was fully integrated with the geo-location mechanism that are associated with blogs; and also pick and choose like the search-engine-image-selector in choosing the best images to make the final panoramic view.

    Summary

    In the end, what would be nice to see is something that not only showed the satellite view, but was a tool that showed me the ground level view of that location.

    There may be other solutions to do this. But an “easy” approach would be to develop a system that indexed Flickr Photographs, and then let users pick a location, and have these ground-level images form a panoramic view of the area of interest.


    LEGAL NOTICE


    Creative Commons License

    This work is licensed under a Creative Commons License.

    You may not copy any of this work to promote a commercial product on any site or medium in the universe.

    If you see this work posted on a commercial site, it violates the creative commons license; and the author does not endorse the commercial product.

    Free to use for non-commercial uses. Link to this original blogspot and cite as .
  • " />