[#GEOT-1651] AbstractGridCoverage2DReader should consider using both overviews and subsampling at the same time

GeoTools

AbstractGridCoverage2DReader should consider using both overviews and subsampling at the same time

Created: 30/Dec/07 10:40 AM Updated: 04/Sep/08 08:04 PM

Component/s:

core coverage

Affects Version/s:

2.5-M0, 2.4-RC0

Fix Version/s:

2.4.6

Description

« Hide

If an image has overviews that do respect a power of 2 scale, picking the right overview is often enough.
Yet, if the overviews do not follow power of 2, or if the last overview is not small enough for the desired resolution, it should be possible to apply subsampling on the chosen overview to further reduce the amount of data to be read.

All

Comments

Change History

Sort Order:

Ascending order - Click to sort in descending order

[ Permlink | « Hide ]

Martin Desruisseaux - 30/Dec/07 10:49 AM

This is the kind of work already done by MosaicImageReader in org.geotools.image.io.mosaic. We could also imagine alternative implementation if the above doesn't suit. But in all case, I believe that this kind of work would be better done at the ImageReader level rather than GridCoverageReader. The AbstractGridCoverage2DReader may need to be replaced by a new class delegating most of its work to the ImageReader. I tend to believe that a single GridCoverageReader implementation wrapping arbitrary ImageReader would be enough for most cases, as opposed to the current approach were we have different AbstractGridCoverage2DReader subclasses for each format. We may be able to handle the various formats strictly at the ImageReader level, providing that we define "standard" Image I/O metadata for geographic informations.

[ Permlink | « Hide ]

Simone Giannecchini - 03/Jan/08 09:40 AM

I have committed on trunk a few lines that should allow what you need Andrea. Please test them and provide some feedback.