Theta V GPS duplicate location ERR

Q&A about the latest versions
Post Reply
User avatar
360Texas
Moderator
Posts: 3684
Joined: Sat Sep 09, 2006 6:06 pm
Location: Fort Worth, Texas USA
Contact:

Theta V GPS position Issue.

We took 3 Theta V images with a Samsung s8 50 feet apart. Found a discrepancy in recording the GPS location.

We used Photoshop CS6 (non Cloud) to open Theta V 3 images and look at IMAGE INFO (RAW tab) . Scrolled down to bottom GPS data info.

Using Pano2vr Mapping this creates an issue.

Found that even though image 1 and 2 were taken 50 feet apart and 4 minutes later... the GPS LAT/ Long were duplication. Here is the GPS data:

From original panorama images :

R0010059_xxxxxxx 1919.jpg Starting GEO position
GPS data
<exif:GPSVersionID>2.3.0.0</exif:GPSVersionID>
<exif:GPSLatitude>32,39.6167N</exif:GPSLatitude>
<exif:GPSLongitude>97,22.2667W</exif:GPSLongitude>
<exif:GPSTimeStamp>2017-12-21T10:19:08Z</exif:GPSTimeStamp>
<exif:GPSImgDirectionRef>T</exif:GPSImgDirectionRef>
<exif:GPSImgDirection>2126/10</exif:GPSImgDirection>
<exif:GPSMapDatum>WGS-84</exif:GPSMapDatum>

Date time stamp 2017-12-21T10:19:08Z

--
R0010061_xxxxxxx 2317 ERR in coordinates even though this image is 50 ft and 4 minutes later from 10059 above
GPS data
<exif:GPSVersionID>2.3.0.0</exif:GPSVersionID>
<exif:GPSLatitude>32,39.6167N</exif:GPSLatitude>
<exif:GPSLongitude>97,22.2667W</exif:GPSLongitude>
<exif:GPSTimeStamp>2017-12-21T10:23:08Z</exif:GPSTimeStamp>
<exif:GPSImgDirectionRef>T</exif:GPSImgDirectionRef>
<exif:GPSImgDirection>758/10</exif:GPSImgDirection>
<exif:GPSMapDatum>WGS-84</exif:GPSMapDatum>

Date time stamp 2017-12-21T10:23:08Z or 4 minutes later than first Image R00100059 above
--

R0010063..dd/mm/yyyy
GPS data
<exif:GPSVersionID>2.3.0.0</exif:GPSVersionID>
<exif:GPSLatitude>32,39.6167N</exif:GPSLatitude>
<exif:GPSLongitude>97,22.25W</exif:GPSLongitude>
<exif:GPSTimeStamp>2017-12-21T10:27:13Z</exif:GPSTimeStamp>
<exif:GPSImgDirectionRef>T</exif:GPSImgDirectionRef>
<exif:GPSImgDirection>200/10</exif:GPSImgDirection>
<exif:GPSMapDatum>WGS-84</exif:GPSMapDatum>

Date time stamp 2017-12-21T10:27:13Z or 8 minutes later than first Image R00100059 above

This creates an issue when GPS data is used in Pano2vr
😎
Dave
Pano2VR Forum Global Moderator
Image
Visit 360texas.com
Don
Posts: 143
Joined: Thu Mar 07, 2013 2:59 pm
Location: Southern California

Thanks for your post Dave! Great to know, and along those lines I have been thinking of this work-around, please update me: Is it now possible in pano2vr from the opposite direction you describe in your post, to use a flat image as a key map, then register the key map's lat/long bounds in pano2vr, then position the nodes (any nodes, not just Theta nodes) manually anywhere on that key map, and behold, achieve an extremely accurate node marker lat/long coordinate? For example, 7 or 8 digits to the Right of the lat/long decimal point, enough to achieve 1/8" accuracy? (And if that happens to be less than the value of a pixel, then rounded to the nearest pixel...as when Tiles are displayed until zoomed in close enough.) I am in a part of the building construction industry where 1/8" accuracy is still the acceptable margin of error.

I have similar dreams for hotspots in panos, but will settle first for precision node markers on a key map. My key map being a high res floor plan. (A floor plan being more precise than an indoor map. An indoor map being more precise than a geographic map.)

Cheers, Don
User avatar
360Texas
Moderator
Posts: 3684
Joined: Sat Sep 09, 2006 6:06 pm
Location: Fort Worth, Texas USA
Contact:

Theta V actually does NOT record the GPS data ON CAMERA. Like the Theta S the V communicates with cell phone to acquire the cellphone's GEO data just after the image is taken and adds the coordinates to the images EXIF header as it is building the panorama. Then it copies the completed image back to the cellphone. So the GEO recorded location is that of the PHONE not where the image was taken. And in this case phone/ camera are within 15' of each other. Given the satellite precision (Civil not military) is probably 30' offset accuracy. The data shown below was with a cellphone connected.

As you can see recorded GPS coordinates are not precision enough to directly get to 1/8th of an inch. Abeit, there are 6 zero's after the decimal point.
Cell and camera connected
Cell and camera connected
thetaVGPS.JPG (102.42 KiB) Viewed 3312 times
Dave
Pano2VR Forum Global Moderator
Image
Visit 360texas.com
User avatar
JCasman
Posts: 1
Joined: Wed Jan 09, 2019 5:10 pm
Location: USA
Contact:

Adding information over a year after the original post. But wanted to point out a GPS plug-in is available for THETA V. It's not a polished plug-in that you download from the THETA Plug-in Store, but it's available for installing and testing. If you're comfortable with working with your THETA V that way, it may allow access to useful functionality. On the theta360.guide site, several Street View people have tried it and report that it's useful. Requires a OTC cable and GPS dongle.

Some initial information here: https://community.theta360.guide/t/usin ... ug-in/3932
User avatar
360Texas
Moderator
Posts: 3684
Joined: Sat Sep 09, 2006 6:06 pm
Location: Fort Worth, Texas USA
Contact:

Wow that was along time ago. Yes, I too soon after learned that the Theta V does have a GPS. Was a bit confusing. I would copy an image from the camera directly to my hard drive then use an app called EXIF Pilot to look through the image EXIF header for gps coordinates... the GPS coordinates would be only LAT 32 39 37.00000. No precision there.

The Theta V while having a GPS its precision is not really useful.

Google Maps APP and smartphone's GPS can actually be quite accurate.

Thanks for looking at this thread and I will take a look at THETA Plug-in Store.
Dave
Pano2VR Forum Global Moderator
Image
Visit 360texas.com
WingmanMedia
Posts: 25
Joined: Mon Jun 03, 2019 12:45 pm
Location: Brisbane, Australia
Contact:

I have just tested a brand new Z1 and it has exactly the same problem. I also have V with the same problem.

I have no idea why Ricoh completelly ignoring such a big bug which makes the whole workflow with publishing their panoramas to GSV super time consuming. If fact it doubles the time required to do it. I recently shoot 330 panos for a client outdoor and it has taken me about 9 hours. Then I spent another 8 hours just to place each panorama on a map. That extra 8 hours could be avoided if they could write location data with more precision.

If there anybody who uses V or Z1 and is happy with location data recorded?

I know about GPS plugin but I would like to see both models recording more precise location by default.
Post Reply