[Geopackage] New Blog Post
Pepijn Van Eeckhoudt
pepijn at vaneeckhoudt.net
Fri Dec 16 13:33:25 EST 2016
There's a difference between not restricting choices and leaving it completely undefined :) The latter is bound to lead to different systems interpreting the data in different ways.
As an example, suppose you naively transcode dted to geopackage, where every source dted sample corresponds to a pixel in a png tile. How is a client supposed to figure out that the samples on the file edges overlap between adjacent tiles for instance?
One client may assume that and convert the elevation samples to a 3D mesh correctly; another one might assume area samples and end up displaying duplicate samples and have every sample at an incorrect location.
If the spec does not describe anything at all regarding interpretation neither of the implementations is either right or wrong.
> On 16 Dec 2016, at 18:41, Jeff Yutzler <jeffy at imagemattersllc.com> wrote:
> We deliberately didn't specify this because we didn't want to mandate one flavor over another. We felt that this could be handled out of band or through the metadata.
>> On Fri, Dec 16, 2016 at 3:27 AM, Pepijn Van Eeckhoudt <pepijn at vaneeckhoudt.net> wrote:
>> Hi Jeff,
>> I might have missed it, but there doesn't seem to be any description of the geographic location of the sample values.
>> Are the pixel values located on grid points or at the grid cell center? Have they been point sampled or are they area averages? etc.
>> This information is pretty critical to know when interpreting the data and interpolating values (evaluating the coverage in topic 6 terminology).
>> Best regards,
>> Pepijn Van Eeckhoudt
>>> On 15 Dec 2016, at 17:42, Jeff Yutzler via Geopackage <geopackage at lists.opengeospatial.org> wrote:
>>> I just published a new blog post regarding the new Elevation Extension:
>>> Jeff Yutzler
>>> Image Matters LLC
>>> Mobile: (703) 981-8753
> Jeff Yutzler
> Image Matters LLC
> Mobile: (703) 981-8753
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Geopackage