Re: [Tagging] Accuracy of survey
While realigning the coastline is possible, they will be surveying for a decade or so just to figure out everything that moved. No, not a decade. While it will take some amount of time for changes to propagate to cartographic products according to their update cycle, the 'figuring out what moved' happens in essentially real time across the major geodetic network, and probably across a month depending on the ephemeris of the JAXA and ESA SARsat http://vldb.gsi.go.jp/sokuchi/sar/index-e.html platforms, although that is probably according to some sort of priority criteria derived from the GNSS data. See Geospatial Information Authority of Japan, １．Continuous observation at the GNSS-based control stations and ４．Synthetic Aperture Radar observation at http://www.gsi.go.jp/ENGLISH/page_e30068.html Coastal survey is longer, because of temporal interval required to interpolate and detect sea level extremes between the phasing of the tides and the satellites, look angles, etc. Michael Patrick Geospatial Analyst http://www.gsi.go.jp/ENGLISH/page_e30068.html ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
According to http://www.dw.de/quake-shifted-japan-by-over-two-meters/a-14909967 it was 2,4 m. 2014-12-30 22:22 GMT+01:00 Rainer Fügenstein r...@oudeis.org: W Ultimate 'accuracy'? You do realise that the tectonic plates are moving? btw: as a result of the Mar.2011 earthquake, japan has moved by at least 5m. how did OSM react? ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
It was over 5 meters in some places along the coast, but only a very small part. Under the ocean, it was 25m. Most of japan stayed put, but the northern section along the coast was stretched a bit wider, but the coast sank about 1m, so with coastal flooding, japan didn't get that much bigger. There was no uniform movement so you'll get a different number depending on what number you like (largest, average for Ojika peninsula, average for North Japan, etc) Anyway you slice it, The Pacific Ocean is a bit smaller now. So Northern Japanese mountains, roads, and farms all have fractionally different dimensions and altitudes now. Everyone talks about the big one, but there were thousands of aftershocks, and a hundred or so very large ones, including ones inland that caused 1m tall fissures, gaps, and trenches to open up all over north japan, severing roads and buildings. All of them have different elevations now and slightly different road alignments. The release of pressure from the big one allowed all of the smaller inland faults to start moving again. Here's a pdf (full of pictures) of a 6.6 a month later that caused 1m uplift and offset in Fukushima. http://www.geerassociation.org/GEER_Post%20EQ%20Reports/Tohoku_Japan_2011/QR4_Preliminary%20Observations%20of%20Surface%20Fault%20Rupture_06.06.11.pdf At the small peninsula closest to the earthquake, Ojika-hanto, the parking lot across the street from the shoreline is the new shoreline. I visited in July 2011. This is completely due to the lowering of the seabed from the quake. https://www.flickr.com/photos/javbw/11091302756/in/set-72157638113676925 The sidewalk is visible(left), but the unclassified road to the far left and the old shoreline, is completely underwater now, as is the pier. While realigning the coastline is possible, they will be surveying for a decade or so just to figure out everything that moved. Javbw Sent from my iPad On Dec 31, 2014, at 9:36 PM, Mateusz Konieczny matkoni...@gmail.com wrote: According to http://www.dw.de/quake-shifted-japan-by-over-two-meters/a-14909967 it was 2,4 m. 2014-12-30 22:22 GMT+01:00 Rainer Fügenstein r...@oudeis.org: W Ultimate 'accuracy'? You do realise that the tectonic plates are moving? btw: as a result of the Mar.2011 earthquake, japan has moved by at least 5m. how did OSM react? ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
Hi Warin and all, I am not sure what you dislike in accuracy. Accuracy is how far the measured mean value is from the actual value ( http://en.wikipedia.org/wiki/Accuracy_and_precision). However we can start calling it a trueness, to follow the ISO definition. If you mean something else, please explain, and I believe it would deserve a page in OSM wiki. Now, I am talking about the absolute trueness. That is, how far a POI is according to the map from its actual position on the planet. No, I don't forget that the planet surface is moving relative to the GPS coordinates. Even more so, there are local surface movements, especially if the survey marker is located, say, on a bridge or close to an excavation site. It should be considered when defining this non-movable POIs. Taking into account the inherent precision of the survey marker position (they are designed to have a well-defined position), it does make sense to have OSM data for them defined better than for all usual elements. At the same time, if you are talking about common use, these POIs are of little interest to normal users. So their specific properties will not disturb anyone. However, some mappers may be in possession of the surveying tools allowing them to have better trueness than possible with a GPS, provided that they have some good reference points. Survey markers are designed just for that. For these mappers, the absolute location of the survey markers is important, and I see no reason to prevent them from having it in OSM. OSM renders distort road widths according to their classification .. that is normal mapping for road navigation. If you wanted air navigation then the actual road width would be better to render, with runways having more emphasis. True. However the underlying data is independent from how a specific renderer represents each element. A street is usually just a line, thus having no width. Cheers, Kotya On Tue, Dec 30, 2014 at 12:50 AM, Warin 61sundow...@gmail.com wrote: On 30/12/2014 6:41 AM, tagging-requ...@openstreetmap.org wrote: Date: Mon, 29 Dec 2014 15:27:23 +0100 From: Kotya Karapetyan kotya.li...@gmail.com kotya.li...@gmail.com To: Rainer Fügenstein r...@oudeis.org r...@oudeis.org, Tag discussion, strategy and related tools tagging@openstreetmap.org tagging@openstreetmap.org Subject: Re: [Tagging] Accuracy of survey Message-ID: cak2dj-whwqajz+0-oxjue9bhn-w1eldcypm4am4xidn2fp5...@mail.gmail.com cak2dj-whwqajz+0-oxjue9bhn-w1eldcypm4am4xidn2fp5...@mail.gmail.com Content-Type: text/plain; charset=utf-8 Since such reference points are quite common, I would support the idea of creating a special tag for them, requiring that they are not moved. However we need a clear consensus on how we define the sufficient accuracy and how the data for such points will be updated. Ultimate 'accuracy'? You do realise that the tectonic plates are moving? So your reference points need to include a date so they can be corrected for the drift. You'll find that data is available for those survey reference points .. together with their precision. Do you want to update these points to maintain their 'accuracy'? How often? Survey reference points are 'quite common' in built up areas ... but not in remote locations. And depending an the age and how precise the survey was will have some effect on their 'accuracy'. One surveor in Australia forget to allow for the temperature effect on this measurement chain back when chains were used. I disagree with the point of view that an accuracy sufficient for consumer GPS devices is sufficient for OSM and therefore there is no problem here. Nobody ever declared that OSM is for smartphone users. We are trying to map the world, and accuracy should be of primary interest for this project. Again the word 'accuracy'. Context 1. I have advised one mapper in their diary that most, if not all, users will be using their data entry with similar equipment to what they have .. so any 'inaccuracy' will also be present for the other users. Thus what they map should represent what is there and should be usable as a map .. considering that the GPS information may be very vague under the tree cover present and the local cliffs etc. Context 2 I will be mapping a track that is covered in a few places .. by an over hanging cliff. As such it is not visible by satellite .. nor will the GPS track be that 'accurate'. So I'll be mapping it from the available information that I have then - a few photos, my track and the satellite image. It will take me about a week to traverse the area. No shops etc. I would rather have the less 'accurate' representation of what is there compared to a blank area. I've plotted one track that goes from one place to another (personal knowledge).. where it is not visible on the satellite view I've plotted it as a straight line.. I know it is not a straight line but it is the best I can do and conveys
Re: [Tagging] Accuracy of survey
W Ultimate 'accuracy'? You do realise that the tectonic plates are moving? btw: as a result of the Mar.2011 earthquake, japan has moved by at least 5m. how did OSM react? ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
2014-12-24 0:21 GMT+01:00 Friedrich Volkmann b...@volki.at: I used estimated_accuracy=* or gps_accuracy=* a couple of times, but I doubt that it prevents other mappers from moving or even deleting them. Some use editors like Potlatch, so they are not aware of tags. Some do thousands of edits, all of which are validator based corrections. They do not ask nor think nor look at tags, except at those reported by the validator. the effects of those semi-mass-edits or other careless following edits must not be feared too much: as long as the original tag is preserved (otherwise it will unlikely be noticed unless it is searched for) other mappers might take a look and see from the history to which coordinates the note belongs. I think notes are a good way of passing particular information about the survey conditions to other mappers. Positional accuracy should not be overestimated, in dense areas it is more important to have good relative positioning (things should relate in the map like they do in the real world, e.g. with regard to left or right side of the road, crossing in the same point or 2 adjacent crossings, angles, line of sight, size relations, parallel vs. not, etc. In these settings you typically won't find a GPS of much use when mapping today in a well mapped urban area. In lower density areas (e.g. countryside, mountain areas) it usually doesn't matter to have cm-precision, 10-15m are more than sufficient, bare some potentially very rare counter examples. Still I can understand that when you use equipment with significant higher or lower precision than average you'd want to have a dedicated tag to formalize entering the presumed precision in a machine readable way. just do it ;-) Cheers, Martin ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
Happy holidays and 2015 everyone! what is needed here is some tag, saying don't touch these coordinates, they've been surveyed with high(est) accuracy. I second this idea. Just recently I discovered that something in this direction already exists: https://wiki.openstreetmap.org/wiki/WikiProject_France/Rep%C3%A8res_G%C3%A9od%C3%A9siques#Permanence_des_rep.C3.A8res Example: https://www.openstreetmap.org/edit#map=23/43.42272/6.76665 However it seems to be France-specific. I don't know if a similar thing exists e.g. for Germany. Since such reference points are quite common, I would support the idea of creating a special tag for them, requiring that they are not moved. However we need a clear consensus on how we define the sufficient accuracy and how the data for such points will be updated. I disagree with the point of view that an accuracy sufficient for consumer GPS devices is sufficient for OSM and therefore there is no problem here. Nobody ever declared that OSM is for smartphone users. We are trying to map the world, and accuracy should be of primary interest for this project. Cheers, Kotya On Thu, Dec 25, 2014 at 7:16 PM, Rainer Fügenstein r...@oudeis.org wrote: TP It was not clear if the OP indeed wants to map pipelines, TP or was just quoting the pipeline expert for his opinion about TP surveying methods. the latter. I'm referring to all nodes, not just pipelines marker. Just used the conversation I had some time ago as an example. W Terms !! W In Metrology (http://en.wikipedia.org/wiki/Metrology) the words W accuracy, error, etc have specific meaning .. please forgive my ignorance - let the experts decide on a proper term to be eventually used as tag. dilution comes to mind, but that's GPS specific, if I'm not mistaken. FV Even if you collect plenty of GPS traces with no systematic error, these FV still cannot beat a theodolite triangulation. when specifying accuracy, the source of the coordinates shouldn't matter. It could be GPS, DGPS, theodolite triangulation, a file provided by officials or companies ... FV I used estimated_accuracy=* or gps_accuracy=* a couple of times, IMHO, that's the way to go. would recommend against gps_*, see above. also, there should be a distinction between estimated and actual accuracy. FV but I doubt FV that it prevents other mappers from moving or even deleting them. Some use FV editors like Potlatch, so they are not aware of tags. Some do thousands of FV edits, all of which are validator based corrections. They do not ask nor FV think nor look at tags, except at those reported by the validator. software evolves; if such a tag is considered useful and widely used, it may eventually be supported by the developers. of course, there'll always be the black sheep ... FV Also, there is no clear line between high and low precision data. Should an FV editor warn when the precision is better than 1m, but ignore a precision of FV 2m? This all depends on the precision of the new data, which the editor does FV not know. for starters, I'd begin with a general warning if the precision of the existing node is less or equal than 2m (thats better than what the average consumer receiver can achieve). to draw a line between high and low precision, this article [1] may be helpful. some GPS receivers show the current precision in meters; GPX files contain HDOP/VDOP/PDOP if provided by the receiver. In theory and if provided, when a GPX file is used as source for nodes, precision could be derived from this information (by whatever means). FV There are no GPS traces for pipeline markes. actually, there are ;-) I just didn't upload mine. but apart from that, pipeline mapping seems to be a few-(wo)men show, therefore it's more likely that pipeline operators may release their (high precision) data [2] before there are enough GPS traces to significantly increase precision via interpolation. cu [1] http://en.wikipedia.org/wiki/Global_Positioning_System (section Augmentation f.) [2] https://wiki.openstreetmap.org/wiki/Talk:Proposed_features/PipelineExtension#status_update.2F1 ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
Kotya Karapetyan wrote on 2014-12-29 15:27: Just recently I discovered that something in this direction already exists: https://wiki.openstreetmap.org/wiki/WikiProject_France/Rep%C3%A8res_G%C3%A9od%C3%A9siques#Permanence_des_rep.C3.A8res Example: https://www.openstreetmap.org/edit#map=23/43.42272/6.76665 Better point directly to the object you refer to: https://www.openstreetmap.org/node/670599510 However it seems to be France-specific. I don't know if a similar thing exists e.g. for Germany. Since such reference points are quite common, I would support the idea of creating a special tag for them, Yes geodetic reference points exist, and as you see 285000 are already mapped in OSM. And they have a tag already, man_made=survey_point https://wiki.openstreetmap.org/wiki/Tag:man_made%3Dsurvey_point requiring that they are not moved. But that is the same for anything in OSM. Reference points should not be moved accidentally (there might be reasons to move them, so you cannot exclude that per se), motorways should not be deleted accidentally. It is a general problem that OSM data are quite vulnerable, and a single tag on a particular object will not solve this. As for reference points, what I find useful to have a human-readable copy of the lat/lon values in a tag, so they can be inspected more easily in case of a movement or recreation (i.e. wiped history). I disagree with the point of view that an accuracy sufficient for consumer GPS devices is sufficient for OSM and therefore there is no problem here. Fine, however nobody had expressed this point of view. You might have misread some postings. Nobody ever declared that OSM is for smartphone users. Oh, OSM _is_ for smartphone users, and any other user of course ;-) We are trying to map the world, and accuracy should be of primary interest for this project. Sure. Tom Cheers, Kotya ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
On 30/12/2014 6:41 AM, tagging-requ...@openstreetmap.org wrote: Date: Mon, 29 Dec 2014 15:27:23 +0100 From: Kotya Karapetyan kotya.li...@gmail.com To: Rainer Fügenstein r...@oudeis.org, Tag discussion, strategy and related tools tagging@openstreetmap.org Subject: Re: [Tagging] Accuracy of survey Message-ID: cak2dj-whwqajz+0-oxjue9bhn-w1eldcypm4am4xidn2fp5...@mail.gmail.com Content-Type: text/plain; charset=utf-8 Since such reference points are quite common, I would support the idea of creating a special tag for them, requiring that they are not moved. However we need a clear consensus on how we define the sufficient accuracy and how the data for such points will be updated. Ultimate 'accuracy'? You do realise that the tectonic plates are moving? So your reference points need to include a date so they can be corrected for the drift. You'll find that data is available for those survey reference points .. together with their precision. Do you want to update these points to maintain their 'accuracy'? How often? Survey reference points are 'quite common' in built up areas ... but not in remote locations. And depending an the age and how precise the survey was will have some effect on their 'accuracy'. One surveor in Australia forget to allow for the temperature effect on this measurement chain back when chains were used. I disagree with the point of view that an accuracy sufficient for consumer GPS devices is sufficient for OSM and therefore there is no problem here. Nobody ever declared that OSM is for smartphone users. We are trying to map the world, and accuracy should be of primary interest for this project. Again the word 'accuracy'. Context 1. I have advised one mapper in their diary that most, if not all, users will be using their data entry with similar equipment to what they have .. so any 'inaccuracy' will also be present for the other users. Thus what they map should represent what is there and should be usable as a map .. considering that the GPS information may be very vague under the tree cover present and the local cliffs etc. Context 2 I will be mapping a track that is covered in a few places .. by an over hanging cliff. As such it is not visible by satellite .. nor will the GPS track be that 'accurate'. So I'll be mapping it from the available information that I have then - a few photos, my track and the satellite image. It will take me about a week to traverse the area. No shops etc. I would rather have the less 'accurate' representation of what is there compared to a blank area. I've plotted one track that goes from one place to another (personal knowledge).. where it is not visible on the satellite view I've plotted it as a straight line.. I know it is not a straight line but it is the best I can do and conveys the information that the track is connected, and being straight in that hilly area also conveys that the information is a guide. I know there is a similar tack a bit north of this track .. but cannot reliably get the entry and exit points .. so have left that off as I view it as unreliable for use. I have come across similar in other areas of the world .. but I found the satellite image had better information - so 'improved' the information. -- OSM primary interest? 1) to be USEFULL. meaning to have information desired by the user sufficient representation and detail to be able to navigate to a desired place. Many usefull maps have distortion - to include more details on particular objects or to simply emphasise to those objects. OSM renders distort road widths according to their classification .. that is normal mapping for road navigation. If you wanted air navigation then the actual road width would be better to render, with runways having more emphasis. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
On 2014-12-23 20:01, christian.pietz...@googlemail.com wrote : the problem with nodes is, that they are easily overssen. It would be nice to have the possibility to show a warning when moving an opject with high accuracy. The best examples are GPS reference points like this: https://www.openstreetmap.org/node/2977364293 Since they have fixed coordinates it would be nice to mark them as such. Not only GPS. Satellite maps and geodesic measurements too. Just a few examples http://overpass-turbo.eu/?Q=[out%3Ajson][timeout%3A25]%3B%0A%28%20node[%22man_made%22%3D%22survey_point%22]%28%7B%7Bbbox%7D%7D%29%3B%0A%29%3B%0A%2F%2F%20print%20results%0Aout%20body%3B%0A%3E%3B%0Aout%20skel%20qt%3BC=48.83986;2.41493;10R. Note the very specific tags. Are we sure that the Are you sure feature should be restricted to coordinates? Cheers André. 2014-12-23 19:10 GMT+01:00 Michael Kugelmann michaelk_...@gmx.de mailto:michaelk_...@gmx.de: Am 23.12.2014 um 17:37 schrieb Rainer Fügenstein: what is needed here is some tag, saying don't touch these coordinates, they've been surveyed with high(est) accuracy. maybe just add a note to the pipeline (note = maped mit GPS with guaranteed accuracy of blahblah). Cheers, Michael. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
the problem with nodes is, that they are easily overssen. It would be nice to have the possibility to show a warning when moving an opject with high accuracy. The best examples are GPS reference points like this: https://www.openstreetmap.org/node/2977364293 Since they have fixed coordinates it would be nice to mark them as such. 2014-12-23 19:10 GMT+01:00 Michael Kugelmann michaelk_...@gmx.de: Am 23.12.2014 um 17:37 schrieb Rainer Fügenstein: what is needed here is some tag, saying don't touch these coordinates, they've been surveyed with high(est) accuracy. maybe just add a note to the pipeline (note = maped mit GPS with guaranteed accuracy of blahblah). Cheers, Michael. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
TP It was not clear if the OP indeed wants to map pipelines, TP or was just quoting the pipeline expert for his opinion about TP surveying methods. the latter. I'm referring to all nodes, not just pipelines marker. Just used the conversation I had some time ago as an example. W Terms !! W In Metrology (http://en.wikipedia.org/wiki/Metrology) the words W accuracy, error, etc have specific meaning .. please forgive my ignorance - let the experts decide on a proper term to be eventually used as tag. dilution comes to mind, but that's GPS specific, if I'm not mistaken. FV Even if you collect plenty of GPS traces with no systematic error, these FV still cannot beat a theodolite triangulation. when specifying accuracy, the source of the coordinates shouldn't matter. It could be GPS, DGPS, theodolite triangulation, a file provided by officials or companies ... FV I used estimated_accuracy=* or gps_accuracy=* a couple of times, IMHO, that's the way to go. would recommend against gps_*, see above. also, there should be a distinction between estimated and actual accuracy. FV but I doubt FV that it prevents other mappers from moving or even deleting them. Some use FV editors like Potlatch, so they are not aware of tags. Some do thousands of FV edits, all of which are validator based corrections. They do not ask nor FV think nor look at tags, except at those reported by the validator. software evolves; if such a tag is considered useful and widely used, it may eventually be supported by the developers. of course, there'll always be the black sheep ... FV Also, there is no clear line between high and low precision data. Should an FV editor warn when the precision is better than 1m, but ignore a precision of FV 2m? This all depends on the precision of the new data, which the editor does FV not know. for starters, I'd begin with a general warning if the precision of the existing node is less or equal than 2m (thats better than what the average consumer receiver can achieve). to draw a line between high and low precision, this article [1] may be helpful. some GPS receivers show the current precision in meters; GPX files contain HDOP/VDOP/PDOP if provided by the receiver. In theory and if provided, when a GPX file is used as source for nodes, precision could be derived from this information (by whatever means). FV There are no GPS traces for pipeline markes. actually, there are ;-) I just didn't upload mine. but apart from that, pipeline mapping seems to be a few-(wo)men show, therefore it's more likely that pipeline operators may release their (high precision) data [2] before there are enough GPS traces to significantly increase precision via interpolation. cu [1] http://en.wikipedia.org/wiki/Global_Positioning_System (section Augmentation f.) [2] https://wiki.openstreetmap.org/wiki/Talk:Proposed_features/PipelineExtension#status_update.2F1 ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
Accuracy is indeed a problem. An early OSM update I made was moving a borderline by 250m and put a devotion site from an arrondissement to another. Since then many corrections of more that 5 m, mainly due to user being unaware of Bing's offset (at close zoom of course, in some places only, poisoned gift) and some editors. I just checked a place where I had spotted a Bing offset before. National aerial photos 2009 and 2012 were offset by 2.8m. Bing was almost in the middle Smart phones have a bad accuracy reputation. I have however bought one based on a "good GPS" user report and, indeed I verified quick fix and immediate 4m accuracy by cloudy weather in a veranda. Former one around 10m and couldn't fix in house. On 2014-12-23 18:49, Malcolm Herring wrote : On 23/12/2014 16:57, Tom Pfeifer wrote: The collection of traces over a longer time creates a cloud of traces which form a Gaussian bell curve, in density, over the ground truth. Except that the position of a node in the DB is the last edited value, not the mean position of all historical values. Yet, I sketched a funny program that analyzes a GPX trace, determines when the car stops (I doubt it could work for foot), takes the mean value of the wandering position until the car moves again, and creates a POI for JOSM to layer. The results were surprising for an alpha 0 pre-release pure hack. After rolling up one's sleeves higher, one could imagine processing several traces, such as ones recorded by buses to determine the coordinates of the stops. I had been surprised by how much the GPX position moves about when the GPS is stopped (compared to when it moves) and I will test that with my new smartphone when better weather returns. Cheers André. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
Oops, I forgot to finish a phrase... I just checked a place where I had spotted a Bing offset before and that disappeared. National aerial photos 2009 and 2012 are offset by 2.8m. Bing is almost in the middle. But that's because of a perspective effect. The national orthos have house walls slanted by about 1.4 m for the house height and 2009 and 2012 are slanted in opposite directions. The roofs are offset but they are at the same position at ground level. That's of course very tricky to compensate if one maps house based on roof contours. Cheers André. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
Friedrich Volkmann wrote on 2014-12-23 23:59: There are no GPS traces for pipeline markes. There are traces for roads and paths only. It was not clear if the OP indeed wants to map pipelines, or was just quoting the pipeline expert for his opinion about surveying methods. And if you walk/drive along the pipeline, you could create a trace, and even repeat over time. These traces can bear a systematic error due to reflections (e.g. under a cliff). Yes of course, and in an urban canyon our multi-trace bell curve would be very flat due to the even larger amount of multipath reception and the position noise it injects. Even if you collect plenty of GPS traces with no systematic error, these still cannot beat a theodolite triangulation. I tend to agree on the practical side, although triangulation has error sources too, and we haven't defined the scale of your experiment, and the number of my traces, for the scientific comparison. ;-) My main argument, if you read it in context, was that crowdsourcing of traces from consumer-grade devices provides a significant statistical advantage over a single trace of such, an effect that the pipeline expert might not have been aware of. tom ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
I would consider that a non-issue as you said, for those reasons: - When it comes to GPS traces on objects that don't move (*), the beauty of crowdsourcing is on our side. The collection of traces over a longer time creates a cloud of traces which form a Gaussian bell curve, in density, over the ground truth. Thus a junction of two road traced again and again is still a good reference point to calibrate aerial imagery. - We are getting access to increasingly better geo-referenced aerial imagery, thus mapping can now use different sources and calibrate between them. The real issue is that in urban areas, lots of object, mostly houses in absence of GPS traces, have been mapped with offset imagery and need to be moved a bit. But this has no implication on tagging. (*) emphasis on fixed objects, since our friends from OpenSeaMap have more difficulties creating such repeatable GPS traces since a ship has no fixed road it would use again and again. tom Rainer Fügenstein wrote on 2014-12-23 17:37: while we are at it, imagine the following situation: mapper A, by means of DGPS, MilStd GPS, crystal ball etc., is able to achieve an accuracy of, say, a few centimeters and uses it to add new nodes (POIs) to OSM. some time later, mapper B with his/her ancestors mechanical GPS device (*), achieving an accuracy of max., say, 15 meters, surveys the same area, figures out that (by his/her point of view) POIs added by mapper A are 15 meters off and corrects their location. what is needed here is some tag, saying don't touch these coordinates, they've been surveyed with high(est) accuracy. I heard this argument from an pipeline expert, noting that marker surveyed with consumer GPS are (for their standards) way off their real location. maybe this is a non-issue after all, if consensus is that consumer GPS accuracy is sufficient enough. cu (*) http://www.kenalder.com/measure/excerpts.htm ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
On 23/12/2014 16:57, Tom Pfeifer wrote: The collection of traces over a longer time creates a cloud of traces which form a Gaussian bell curve, in density, over the ground truth. Except that the position of a node in the DB is the last edited value, not the mean position of all historical values. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
Am 23.12.2014 um 17:37 schrieb Rainer Fügenstein: what is needed here is some tag, saying don't touch these coordinates, they've been surveyed with high(est) accuracy. maybe just add a note to the pipeline (note = maped mit GPS with guaranteed accuracy of blahblah). Cheers, Michael. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
MK maybe just add a note to the pipeline (note = maped mit GPS with MK guaranteed accuracy of blahblah). I'm rather thinking of something machine-readable, enabling the editor to warn the user in case he/she is about to change high precision data. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
On 23.12.2014 17:57, Tom Pfeifer wrote: I would consider that a non-issue as you said, for those reasons: - When it comes to GPS traces on objects that don't move (*), the beauty of crowdsourcing is on our side. The collection of traces over a longer time creates a cloud of traces which form a Gaussian bell curve, in density, over the ground truth. Thus a junction of two road traced again and again is still a good reference point to calibrate aerial imagery. There are no GPS traces for pipeline markes. There are traces for roads and paths only. These traces can bear a systematic error due to reflections (e.g. under a cliff). Even if you collect plenty of GPS traces with no systematic error, these still cannot beat a theodolite triangulation. - We are getting access to increasingly better geo-referenced aerial imagery, thus mapping can now use different sources and calibrate between them. In places where GPS is most inaccurate, e.g. in a gorge covered by woods, aerial images are inaccurate too, and most of the ground details are not visible. -- Friedrich K. Volkmann http://www.volki.at/ Adr.: Davidgasse 76-80/14/10, 1100 Wien, Austria ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
On 23.12.2014 17:37, Rainer Fügenstein wrote: mapper A, by means of DGPS, MilStd GPS, crystal ball etc., is able to achieve an accuracy of, say, a few centimeters and uses it to add new nodes (POIs) to OSM. some time later, mapper B with his/her ancestors mechanical GPS device (*), achieving an accuracy of max., say, 15 meters, surveys the same area, figures out that (by his/her point of view) POIs added by mapper A are 15 meters off and corrects their location. what is needed here is some tag, saying don't touch these coordinates, they've been surveyed with high(est) accuracy. I used estimated_accuracy=* or gps_accuracy=* a couple of times, but I doubt that it prevents other mappers from moving or even deleting them. Some use editors like Potlatch, so they are not aware of tags. Some do thousands of edits, all of which are validator based corrections. They do not ask nor think nor look at tags, except at those reported by the validator. -- Friedrich K. Volkmann http://www.volki.at/ Adr.: Davidgasse 76-80/14/10, 1100 Wien, Austria ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Accuracy of survey
On 24/12/2014 11:29 AM, tagging-requ...@openstreetmap.org wrote: Date: Tue, 23 Dec 2014 17:37:34 +0100 From: Rainer Fügenstein r...@oudeis.org To: Tag discussion, strategy and related tools tagging@openstreetmap.org Subject: [Tagging] Accuracy of survey Message-ID: 811143140.20141223173...@oudeis.org Content-Type: text/plain; charset=ISO-8859-1 mapper A, by means of DGPS, MilStd GPS, crystal ball etc., is able to achieve an accuracy of, say, a few centimeters and uses it to add new nodes (POIs) to OSM. some time later, mapper B with his/her ancestors mechanical GPS device (*), achieving an accuracy of max., say, 15 meters, surveys the same area, figures out that (by his/her point of view) POIs added by mapper A are 15 meters off and corrects their location. what is needed here is some tag, saying don't touch these coordinates, they've been surveyed with high(est) accuracy. I heard this argument from an pipeline expert, noting that marker surveyed with consumer GPS are (for their standards) way off their real location. maybe this is a non-issue after all, if consensus is that consumer GPS accuracy is sufficient enough. cu Terms !! In Metrology (http://en.wikipedia.org/wiki/Metrology) the words accuracy, error, etc have specific meaning .. The common term 'accuracy' or 'error' may mean a normal Gaussian distribution taken at 1 sigma or it could mean a rectangular distribution taken at 100% coverage ... probably the semi range. Does it matter? To the normal user .. no. They too have a consumer grade GPS .. and that too will have errors so an offset will be expected. The professional surveyor may use OSM . But not as a source for performing a survey, but as an indication. They should have access to professionally defined points and boundaries, past professional surveys. So I'd say .. No it does not matter. Provided the relationship to the adjacent objects is a reasonable representation of what is there so recognition and navigation is possible. Best tag? I've seen a node with a note 'move to +32.14342 151.345345' (the numbers are fictitious). I did as it said and that conformed much better to the Bing sat image than before it was moved. So I left it there. Troubling no source was stated, that would have been nice. As would have been an 'accuracy' statement. If I have seen a statement of 'uncertainty' with a 'coverage' factor I'd be surprised! Best survey? Ideally ? Several GPS traces done on separate days/weeks/months so as different satellite constellations are used. Reference could also be made to any professional survey marks (usually administered by the government ,these will have known locations and 'accuracies'). These can then be applied to a satellite image but there are significant shifts depending on topography and parallax errors, so caution is required in there application over any significant area. If you don't want something changed .. add a note and say why ... and inculcate the source by adding the source tag Things that are very well documented tend to be left alone. ___ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
