On Dec 28, 2007 4:12 PM, Martin Desruisseaux <[EMAIL PROTECTED]> wrote: > I just wonder... What is the time frame for your mosaic work? I'm working on > "org.geotools.image.mosaic" at exactly the same time. It is deconnected from > imagemosaic for now, because: > > * imagemosaic works at AbstractGridCoverage2DReader level > * MosaicImageReader works at ImageReader level, which is a lower level. > > MosaicImageReader should be faster and consume less memory because it avoid > the > creation of temporary images, and also avoid completly JAI's "Mosaic" > operation > since the mosaic is constructed at reading time instead than assembling it > after > we loaded the tiles. >
I am curious about this statement because it seems a bit optimistic. Can you be a little more specific? How do you assemble them if you are not using the mosaic? Are you designin with in mind using deferred loading via JAI ImageRead or directly the ImageRead as you showed me for SeaGis? If so I doubt that for large coverage you would use less memory unless you are sure you read a really small part of it. > It could be used as the backend for "imagemosaic", but we will need to > redesign > AbstractGridCoverage2DReader before doing so, which I can do only in a few > months. > > Doing so may help to solve some of the issues you reported like the mismatch > between AffineTransforms and the tile layout - maybe because of my work in > referencing, and tend to be very careful with AffineTransforms. > the tile dimensions are passed forward along JAI chains and what he reported > is exactly that. > However if you need "imagemosaic" now, I can not help. It will take me months > before I can reach some GridCoverageReader level. > Having some hints about what you are trying to do could help people with deciding if it's worth to wait a bit and/or join your effort, because it might turn out that your design is good for you but not for other use cases that you were not aware of or simply did not consider, and I see a lot of added value in cooperation at the beginning. Simone. > Martin > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2005. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Geotools-devel mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/geotools-devel > -- ------------------------------------------------------- Eng. Simone Giannecchini President /CEO GeoSolutions S.A.S. Via Carignoni 51 55041 Camaiore (LU) Italy phone: +39 0584983027 fax: +39 0584983027 mob: +39 333 8128928 http://www.geo-solutions.it ------------------------------------------------------- ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Geotools-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/geotools-devel
