Hi again, Thanks for the feedback on building traces. The consensus seems to be for a JOSM plugin while others saying all surveying should be done on the ground. Surveys are only practical where there is permissive or public access to buildings, but the majority of buildings are inaccessible to the public. Therefore, we need a source besides us doing the survey, assuming we want building outlines at all. I think building outlines would be useful for navigation, planning, analysis and many other uses of the map data.
In response to the comment "no imports ever", I would point out that building imports is a completely different situation than that of public roads. A glance at OS Street View suggests that about 99% of buildings are not publically accessible. If we use only manual surveying, we can only achieve coverage of about 1%. I don't think that is satisfactory. Imports are therefore very much appropriate for buildings. The point that buildings may need local knowledge for a high quality import is a better point against a mass import. But again, local knowledge only really applies to the minority of buildings that are accessible. The vast majority of the data can't be improved by local knowledge. I suggest the minority of buildings that can be improved can be done once they have been imported into the OSM database. Regarding the technical difficulty, it remains to be seen if automatic tracing is possible and of sufficient quality. But it is presumptuous to assume it is impossible at this stage. I observed there are slight artifacts in the rectified tiles with are not present in the original set (link below). Notice the "a" in stag is distorted and also some of the letters in "Court". An attempt to automatically trace those tiles would end up with noticeable glitches. Until that stuff gets ironed out, a JOSM plugin, which would presumably grab these tiles via WMS, would have poor automatic tracing performance. http://grant.dev.openstreetmap.org/os-streetview-tiles/17/65322/43740.png Also the rectified tiles seem to be not the highest zoom level available? Having the highest resolution makes automatic tracing much more accurate. There are various service roads that are on the OS opendata site (search for SU986502) that are not in the rectified tiles: http://grant.dev.openstreetmap.org/os-streetview-tiles/17/65322/43739.png One option is to automatically trace objects from the original images and then transform the polygons into WGS84. The main thing I am missing is an practical (and open) OSTN02 implementation (in python). I will investigate this and perhaps trial it in my local area. One question, is this data set going to be maintained by OS in the future? And don't worry, I won't be doing a mass import without a great deal of work and discussion. TimSC _______________________________________________ Talk-GB mailing list [email protected] http://lists.openstreetmap.org/listinfo/talk-gb
