USDA Forest Service, Remote Sensing Applications Center, FSWeb: http://fsweb.rsac.fs.fed.us WWW: http://www.fs.fed.us/eng/rsac/ Optimizing Cache Levels for High Resolution Aerial Imagery James Monty RedCastle Resources, working onsite at Remote Sensing Applications Center Salt Lake City, UT
29
Embed
Optimizing Cache Levels for High Resolution Aerial Imagery€¦ · USDA Forest Service, Remote Sensing Applications Center, . Imagery in the Forest Service . Each National Forest
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
USDA Forest Service, Remote Sensing Applications Center, FSWeb: http://fsweb.rsac.fs.fed.us
WWW: http://www.fs.fed.us/eng/rsac/
Optimizing Cache Levels for High Resolution Aerial Imagery
James Monty RedCastle Resources, working onsite at
Remote Sensing Applications Center Salt Lake City, UT
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Outline
of Imagery in the forest
Ch
al
le
ng
es
tak
ea
wa
ys
I m a g e c a c h e s
How we solved the issue
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery in the Forest Service
Each National Forest has been photographed many times, some as far back as the 1930’s and ‘40’s.
1930 1950 1970 1990 2010 Future
Widespread use of Natural Color & CIR Aerial Photos and 4-band Digital in USDA Forest Service
Presenter
Presentation Notes
Snapshots in time!
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery in the Forest Service
Presenter
Presentation Notes
Rich history of frame photography, with some oblique from the 30’s (top) through the 70’s (bottom) and 80’s. Previously our forests would share imagery on local drives. Finding that for many projects there is a need to more easily share imagery without having it on a number of computers. Also a push within the Forest Service to centralize data.
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery in the Forest Service
Presenter
Presentation Notes
Some forests finding use in commercial sat Lidar gaining momentum as use knowledge expands
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery Use in the Forest Service
• Capture current state
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery Use in the Forest Service
• Capture current state • Planning
Presenter
Presentation Notes
Mix use of NAIP and RP
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery Use in the Forest Service
• Capture current state • Planning • Harvest
Presenter
Presentation Notes
Particularly AK with their own platform – used in conjunction with ground work to do estimates
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery Use in the Forest Service
• Capture current state • Planning • Harvest • Vector updates
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery Use in the Forest Service
• Capture current state • Planning • Harvest • Vector updates • Change analysis
Presenter
Presentation Notes
Use NAIP pair of an fire location
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery Use in the Forest Service
• Capture current state • Planning • Harvest • Vector updates • Change analysis • Image derived products
Presenter
Presentation Notes
FVMs, RDW, atlas
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery Use in the Forest Service
forest-atlas.fs.fed.us
Presenter
Presentation Notes
FVMs, RDW, atlas
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Imagery Use in the Forest Service
Presenter
Presentation Notes
Image server use For me: primarily image services Others are creating map and feature services, but most are uncached
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Challenge in High Res Imagery
Typical user views:
Presenter
Presentation Notes
Our users like to view at ‘insect scale’ Three images: Small scale Large scale Insect scale
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Image Server caches
• Analogous to reduced resolution layers.
• Image tiles delivered to user’s computer.
• Improve redraw rate in revisited locations.
Graphic :ESRI training
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Image Server caches
L19 1:1,128
L18 1:2,257
L17 1:4,514
L16 1:9,028
L15 1:18,056
1:13,542
1:6,771
1:3,386
1:1,693
Presenter
Presentation Notes
Default values – show levels on right hand side graphic so it can be used through the next few slides. Discuss the range that a cache level appears, and how the undesired affect occurs near the upper end.
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
The issue For resource photography, service caches have poor appearance at large scales
Cached Dynamic
6 inch imagery; L17 cache
L19 1:1,128
L18 1:2,257
L17 1:4,514
L16 1:9,028
L15 1:18,056
1:13,542
1:6,771
1:3,386
1:1,693
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
The issue
30 cm imagery; L17/L18 cache break
1:1700 1:1680
L19 1:1,128
L18 1:2,257
L17 1:4,514
L16 1:9,028
L15 1:18,056
1:13,542
1:6,771
1:3,386
1:1,693
For resource photography, service caches have poor appearance at large scales
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Potential Solutions
Just throw more cache levels at it - Most caches go to L17 - Add L18 and L19 Create our own pixel sizes/scales at finer cache levels
L19 1:1,128
L18 1:2,257
L17 1:4,514
L16 1:9,028
L15 1:18,056
1:13,542
1:6,771
1:3,386
1:1,693
Presenter
Presentation Notes
Adding caches does not address the issue of blurry trees near the level breaks Transition: how do we decide on new levels??
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Developing New Cache Levels
Default scale levels are factored at 0.5
Looked into a variety of new factors from 0.6 to 0.8
Level Scale (1:)
Pixel (m)
15 18056 4.77
16 9028 2.39
17 4514 1.19
18 2257 0.59
19 1128 0.29
20 --- ---
21 --- ---
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Developing New Cache Levels
Current levels (0.5)
Level Scale (1:)
Pixel (m)
15 18056 4.77
16 9028 2.39
17 4514 1.19
18 2257 0.59
19 1128 0.29
20 --- ---
21 --- ---
Level Scale (1:)
Pixel (m)
15 18056 4.77
16 12639 3.34
17 8847 2.34
18 6193 1.64
19 4335 1.15
20 3035 0.80
21 2124 0.56
Option 1 (0.700)
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Developing New Cache Levels
Current levels (0.5)
Level Scale (1:)
Pixel (m)
15 18056 4.77
16 9028 2.39
17 4514 1.19
18 2257 0.59
19 1128 0.29
20 --- ---
21 --- ---
Level Scale (1:)
Pixel (m)
15 18056 4.77
16 11736 3.11
17 7629 2.02
18 4959 1.31
19 3223 0.85
20 2095 0.55
21 1362 0.36
Option 2 (0.65)
Presenter
Presentation Notes
Since most of our resource photography is flown to provide a 25 – 30 cm pixel, these seemed like the ideal level for the finest cache.
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Developing New Cache Levels
Improvement at the break
L19 1:3,574
L18 1:5,358
L17 1:8,033
L16 1:12,043
L15 1:18,056
1:14,896
1:9,683
1:6,294
1:4,091
1:2,659
1:1,728
L20 1:2,383
L21 1:1,589
30 cm imagery; L20/L21 cache break
1:1720 1:1740
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Developing New Cache Levels
L19 1:3,574
L18 1:5,358
L17 1:8,033
L16 1:12,043
L15 1:18,056
1:14,896
1:9,683
1:6,294
1:4,091
1:2,659
1:1,728
L20 1:2,383
L21 1:1,589
Easier to find the insects
Presenter
Presentation Notes
Our users can now see the trees in the forest more easily, and also find their insects. Mention that the increase required just under twice as much storage
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Takeaways
» Finding that ArcGIS for Server is more flexible than the older server version
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Takeaways
» Finding that ArcGIS for Server is more flexible than the older server version
» Don’t just accept the defaults; experiment
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Takeaways
» Finding that ArcGIS for Server is more flexible than the older server version
» Don’t just accept the defaults; experiment » That said, only frustration: cache levels
and AGOL
USDA Forest Service, Remote Sensing Applications Center, http://fsweb.rsac.fs.fed.us
Takeaways
» Finding that ArcGIS for Server is more flexible than the older server version
» Don’t just accept the defaults; experiment » That said, only frustration: cache levels
and AGOL » Questions or issues? Wondering about
what others have tried or seen? Hit up the user community. (GeoNet)
Presenter
Presentation Notes
Questions or issues? Wondering about what others have tried or seen? Hit up the user community. (GeoNet)
USDA Forest Service, Remote Sensing Applications Center, FSWeb: http://fsweb.rsac.fs.fed.us
WWW: http://www.fs.fed.us/eng/rsac/
Comments/Questions?
Optimizing Cache Levels for High Resolution Aerial Imagery