Discussion Topic |
|
This thread has been locked |
phylp
Trad climber
Millbrae, CA
|
|
|
Aug 17, 2013 - 11:31pm PT
|
the top couple inches is soft....down below that, bulletproof. A two-fold danger....balling up in the crampons which could EASILY cause a fall
A report with exactly this scenario in last year's Accidents in NAM. Very experienced climber...
|
|
Delhi Dog
climber
Good Question...
|
|
|
Aug 18, 2013 - 12:35am PT
|
Good work mattyj.
That's been bugging me for awhile now.
|
|
Klimmer
Mountain climber
|
|
|
Aug 18, 2013 - 01:59am PT
|
Tom,
You can make Hi-Res 3D stereopairs ... it would really help. Nothing like seeing depth of field to determine what you're looking at.
Zoom in and view with a pokescope or similar viewer:
http://www.pokescope.com/pokescopeinfo.html
|
|
TomCochrane
Trad climber
Santa Cruz Mountains and Monterey Bay
|
|
|
Aug 18, 2013 - 02:20am PT
|
|
|
TomCochrane
Trad climber
Santa Cruz Mountains and Monterey Bay
|
|
|
Aug 18, 2013 - 02:49am PT
|
|
|
SalNichols
Big Wall climber
Richmond, CA
|
|
|
Aug 18, 2013 - 04:56am PT
|
Klimmer, you do know that two copies of the same image DO NOT make a stereo image...right? Unless Tom was shooting with two side by side cameras, you're not going to get the parallax needed for depth perception. Great idea though.
|
|
Klimmer
Mountain climber
|
|
|
Aug 18, 2013 - 09:45am PT
|
Aug 18, 2013 - 01:56am PT
Klimmer, you do know that two copies of the same image DO NOT make a stereo image...right? Unless Tom was shooting with two side by side cameras, you're not going to get the parallax needed for depth perception. Great idea though.
Wrong. Aerial remote sensing to make 3D images only requires a certain percent % overlap in sequential imagery of a flight line. The above images posted from Tom's work are in 3D for everything overlapping. They were a perfect example.
Grab a textbook on photogrammetry. A whole science is developed around this technique. NASA and all the Earth Sciences use this technique for mapping, making topo maps, looking at the surface of the Moon, Mars. etc etc.
A lot of Hi-Res up-close 3D images can be made from Tom's original imagery with sequential frames of overlap.
|
|
monolith
climber
SF bay area
|
|
|
Aug 18, 2013 - 10:25am PT
|
Klimmer can make bad images that fool him, like believing there is an Ark on the moon or his claim that the plane going east over LA was actually a missile going the opposite direction.
[unnecessary link deleted]
Klimmer has had very poor results with his 3D obsession, so I wouldn't put any hope in his work.
|
|
TomCochrane
Trad climber
Santa Cruz Mountains and Monterey Bay
|
|
|
Aug 18, 2013 - 11:17am PT
|
early this morning finished the task taking many hours to download the RED RAW files to sets of five Blue Ray disks
be sure that i know who has the facilities to view these...i.e. Final Cut Pro, RED CineX, Avid, or Aperture...
seconding Cragman, please lay off on the discrimination against Klimmer, he makes some valuable contributions that i appreciate
i agree that two subsequent frames taken in flight past the same area can be aligned to a center point and processed into 3D pairs, providing a valuable search technique that i would like to learn to do
this is a technique that was used extensively by a photo analysis team in England during WWII using Spitfire fighters modified with large format cameras flying some of the most challenging sorties of the war to identify the V1 and V2 launch sites and target them for the bombers just before D-day
|
|
tdg119
Social climber
Northampton, PA
|
|
|
Aug 18, 2013 - 11:24am PT
|
For Cragman:
"I think that is the glacier we visited. We traversed the northwest shore of Cecile lake about 100 ft above the lake. At the end of the lake went slightly left up rock rib and gully . Some class 4 climbing to get up to snow/ glacier. Then diagonally traversed the glacier from climbers left to climbers right up to north notch from notch we could look directly down to iceberg lake. Then went down notch on west side and searched gully system to the left as we were descending. West side of notch was very loose and dangerous but no signs of recent passage. Generally retraced our steps back up and thru notch and down to Cecile an then minaret lake . If he wants to talk PM out phone number.
Thank you
Jill"
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Aug 18, 2013 - 11:34am PT
|
monolith makes some important points regarding the difficulties of reconstructing a 3D representation from a series of 2D shots...
I have read photogrammetry texts, and have worked extensively in understanding this sort of reconstruction.
What is important to remember is that the reconstructions are NOT unique, and especially where the attributes of the camera are unknown and/or uncalibrated, and the positions and times and attitudes of the camera are not known precisely and accurately.
To perform the reconstructions and eliminate these unknowns with the data is a very complex process which would take more time then we have, and skills and computer power which we also do not have (collectively).
The subsequent 3D reconstructions may provide clues which are questionable, and lead searchers into harms way. As such, it is irresponsible to advocate that the simple minded 3D reconstructions be acted upon in a SAR setting. The last thing anyone wants is to have any more tragedy heaped onto this event.
It is important to remember that we can get an apparent 3D reconstruction easily, but when we are using it to provide this sort of information, those easy reconstructions may not correspond to ground truth.
|
|
TomCochrane
Trad climber
Santa Cruz Mountains and Monterey Bay
|
|
|
Aug 18, 2013 - 11:51am PT
|
i have moved the aerial imagery to a separate thread 'Ritter Range Aerial Search Imagery'
my hope is these may prove useful for area familiarization and coordinating searches
these seem to be the highest res accepted by SuperTopo, but are not adequate to spot a person
we are able to zoom in substantially on the original images using tools such as RED CineX or Aperture
we ave a few copies for people who are set up with Aperture or another similar program of your choice
|
|
10b4me
Ice climber
Wishes-He-Was-In-Arizona
|
|
|
Aug 18, 2013 - 11:53am PT
|
Tom, can you refresh my memory? Where was the photo taken that looked like someone was waving at you?
Thanks
|
|
TomCochrane
Trad climber
Santa Cruz Mountains and Monterey Bay
|
|
|
Aug 18, 2013 - 11:57am PT
|
that is the northeast glacier on Banner Peak, near the lower margin of the glacier
that glacier might be interesting for snow/ice climbing, but does not provide a summit route unless you are looking for a challenging rock climb
we made multiple passes of that area, and the interesting anomaly only appears in one pass (there are normally lots of people wandering around up there)
however this is typical of why we have been thinking about going up for another run
we would also like to try a faster frame rate and shutter speed
on a calm day we can also approach terrain more closely...and/or we could handhold the RED camera from a helicopter
it is actually easier and more economical on fuel for us to go to the range directly from our home airport so we don't have to add in the 9000' descent and climb back out of Mammoth Airport in the Owens Valley
we have been spending many hours looking at images that you could only see at a glance while up in the air
|
|
10b4me
Ice climber
Wishes-He-Was-In-Arizona
|
|
|
Aug 18, 2013 - 12:07pm PT
|
Ok. Yeah, that's the one I mentioned up thread about an improbable place to go.
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Aug 18, 2013 - 12:10pm PT
|
you didn't need 3D to do that, Ron... just high resolution 2D...
and if the camera is set up to be a stereoscopic camera, taking two simultaneous images with well defined separation and well calibrated cameras... you're way ahead of the game.
In this case, there is one uncalibrated camera, pointed by hand on a moving platform of unknown position and velocity...
...how is that "easy"?
and on the basis of that information, you're going to claim enough "truth" to send people into some of those places?
You may be comfortable doing that, I am very much not. Let's be responsible here, you haven't the faintest idea of what it takes to provide the sort of information from 3D reconstructions that would justify the risks.
|
|
TomCochrane
Trad climber
Santa Cruz Mountains and Monterey Bay
|
|
|
Aug 18, 2013 - 12:38pm PT
|
the WWII photo analysts were using a single camera out of the belly of a Spitfire
i have to go offline for a while to shuffle equipment, you can still contact me by email
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Aug 18, 2013 - 12:48pm PT
|
I do know something about USFS aerial photography, enough to know that it is a process, not just a random set of images shot by an unknown camera... please Ron.
the speed of a plane being, say 120 miles per hour is 176 feet per second
I don't know the framing rate of the camera, I believe this camera has a variable frame rate from 1 to 120 frames per second...
then the minimum interframe displacement is: 176/120 = 1.47 feet
If the distance to the object is say, 1000 feet then the angular separation of those two frames is:
1.47/1000 = 1.47 milliradians
with 4000 pixel resolution each pixel has a 1.47 mrad/4000 = 36 microradian angular resolution.
Note that such a close spacing of images of adjacent frames will give a poor distance to object reconstruction, you would like the frames to be much more separated (you need parallax).
In addition, the angular velocity of the aircraft is something like 176/1000 = 0.176 radians per second...
This camera may also have a "rolling shutter" which means that the rows of pixels are read out at different times, which means the image is composed of rows taken at different angles, further complicating the reconstruction of even the 2D images...
It depends on the details of the camera...
finally, the aircraft position and attitude may be known for its trajectory, this has to be cross calibrated with the cameras position and attitude for its trajectory (mostly an issue of getting the two time bases calibrated). This is not inconsequential, and much of the information is missing (I don't believe the camera records its position and attitude internally).
If you had the intention of doing a 3D reconstruction you'd take care of these issues before you shot the images... and made sure that the required information exists and that the necessary calibrations are performed, and that the flight trajectory planned to obtain the optimal images for the reconstruction you are attempting.
If you all want to claim that the 3D information is sufficient to send SAR into the areas based on that information you should think more carefully than giving irrelevant anecdotal justification.
|
|
crankster
Trad climber
South Lake Tahoe, CA
|
|
|
Aug 18, 2013 - 12:52pm PT
|
I'm sure some have interesting backgrounds in aerial SAR, but can we refocus this to the search for Matt?
|
|
10b4me
Ice climber
Wishes-He-Was-In-Arizona
|
|
|
Aug 18, 2013 - 12:55pm PT
|
Has anyone asked at Mammoth Mountaineering if they saw, or talked to Matt?
One of the first things I do, if i am in a new area, is to hit up the local climbing shop for beta.
|
|
|
SuperTopo on the Web
|
Let us know!