<div dir="ltr"><div><div><div><div><div><div><div><div><div><div>Hi Michelle,<br><br></div>JOSM is indeed an editor of Openstreetmap data written in JAVA.<br><br></div>Before Mapillary was around, it was already possible to work with pictures in JOSM. These pictures were either geocoded already or one could use a GPX track recorded during the survey, then combine the pictures based on their time stamps.<br><br></div>This means a lot of the functionality we need is already available in JOSM.<br><br></div>Now that Mapillary created apps for all kinds of smartphones, things have changed considerably. On the one hand it became easier to make the pictures, on the other hand, they are already geocoded and if the smartphone has a compass they also include in which direction they were taken.<br><br></div>It also means that it became a lot easier to use pictures taken by other people.<br><br></div>In iD this is already possible, but one has to figure out how to switch on that layer. I don't think many people do figure out how to do it and following a sequence is not very practical either. That's not relevant to this proposal though...<br><br></div>Concerning the issue with thousands of pictures. A solution might be to only download thumbnails to the cache when zoomed in far enough. Another solution might be a dialog window asking whether the user is sure to want that download in case there would be many pictures to download. This might be tricky, as the download happens in a background thread. I'm not sure of this and it will require some experimentation to figure out what works.<br><br></div>The visualisation is indeed icons on the map background indicating where the pictures were taken.<br><br></div>Cheers,<br><br></div>Jo<br><div><div><div><div><div><div><div><div><div><div><div><br></div></div></div></div></div></div></div></div></div></div></div></div><div class="gmail_extra"><br><div class="gmail_quote">2015-03-24 10:20 GMT+01:00 Michelle W.X.Yang <span dir="ltr"><<a href="mailto:wx.michelle.yang@gmail.com" target="_blank">wx.michelle.yang@gmail.com</a>></span>:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div>Hey Jo,<br><br></div><div>Thanks for your email. Please see the inserted lines.<br></div><br><br><div class="gmail_extra"><div class="gmail_quote"><span class="">On Mon, Mar 23, 2015 at 2:07 AM, Jo <span dir="ltr"><<a href="mailto:winfixit@gmail.com" target="_blank">winfixit@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div>Hi Michelle,<br><br></div>Sorry for the late response. I missed your message somehow.<br><br></div>Did you already try to make some pictures and a GPX track, then use JOSM? Load the GPX track, click right mouse button, import pictures.<br><br></div>Then click on one of those pictures. A new window will open, where one can browse from one picture to the next/previous.<br></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></blockquote><div><br></div></span><div>Yes, I have tried this. It's really interesting. So may I understand that JOSM is a tool for editing the OpenStreetMap? And we use the images downloaded from Mapillary as well as photos taken by users of JOSM (ourselves) to improve the OpenStreetMap (generate the street views), is that correct? And the input of JOSM is GPX tracks, and once a GPX track is loaded, the Mapilary plugin (the output of our assignment) needs to download images from Mapillary.com for the area represented by the input GPX track, and store them in a cache. This function can be activated by pressing a button or automatically.<br></div><span class=""><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><br></div>This is more or less what the plugin should provide, but instead with pictures which come from Mapillary.com for the area in view (when zoomed in far enough).<br><br></div>The other 'exercise' you could do is install the Mapillary app on a smartphone, create an account and go out to make some pictures. Upload them and go to view the area on Mapillary.com.<br></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></blockquote><div><br></div></span><div>Yes, I have tried this too. In my side, it seems that iD hasn't provided the 'Mapillary layer' yet, and I can't edit the uploaded pictures.<br></div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><br></div>Click one of the pictures, then the Information icon, then edit with iD. Now you get an editor in your browser. You have to manually switch on the 'Mapillary layer'. (It would be nice if iD could do that for you, but alas).<br><br></div>Once the layer is activated you can click on the pictures to view them.<br><br></div>Unfortunately there is no way to follow a sequence with iD.<br><br></div>The other disadvantage is that it's not easy to use custom aerial imagery and a lot of other handy tools we have in JOSM are missing as well.<br><br></div>So enabling download of Mapillary pictures in the current viewport to a cache<br></div></div></div></div></div></div></div></div></div></div></div></div></div></blockquote></span><div>I got it. But there is a potential issue that if there are thousands of pictures related to the current viewport, maybe we should only choose part of them to visualize. <br></div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div><div><div><div><div><div><div><div><div><div></div>Then visualizing them on the background, showing their 'orientation'<br></div></div></div></div></div></div></div></div></div></div></div></div></blockquote></span><div>May I understand the 'visualizing' as 1) displaying the trace when taking the sequence of photos downloaded from Mapillary, 2) displaying a thumbnail or a drop symbol of each photo downloaded from Mapillary; 3) by double clicking a thumbnail/drop symbol, a window will popup to display this photo with full resolution and user can edit it ? <br></div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div><div><div><div><div><div><div><div><div></div>Then providing the possibility to look at them one picture at a time<br></div>Then having some handy controls to get from one to the next<br></div></div></div></div></div></div></div></div></div></div></blockquote><div> </div></span><div>Intuitively, We can provide arrows in the popup window mentioned above for users to navigate these downloaded photos.<br></div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div><div><div><div><div><div><div><br></div>Is more or less the basis of the assignment.<br><br></div>When these things have become possible, the next step may be the following:<br><br></div>Allow the possibility to update the geolocation of the picture (GPS devices are not always very accurate). <br></div></div></div></div></div></div></div></blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div><div><div><div></div>Allow the possibility to update the 'orientation' of the picture (the internal compass in consumer phones is on the slow side)<br></div><div>Updating the orientation of the picture could be done with +/- or up/down left/right arrows on the keyboard. Another possibility might be the use of a draggable 'helper' node, then calculate the directionality.<br></div></div></div></div></div></div></blockquote><div><br></div></span><div>OK. Once the main frame of this mapillary layer is developed, I believe the above functionalities will not be difficult to be implemented.<br></div><span class=""><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div><div><div></div><div><br></div>Keep track of these changes in a sort of changeset and allow this information to be fed back to Mapillary using their API. The team at Mapillary is very approachable and they can be contacted in case of problems. Authentication happens over oAuth.<br><br><br></div>The next step after that, may be to allow blurring/unblurring as well.<br><br></div>And if time allows, it may be possible to allow updating of recognised traffic signs and possibly an interaction with the roadsigns plugin. But that is low priority at best.<br></div></div></div></blockquote></span><div><br>OK. <br></div><span class=""><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><br></div>Please let us know if you have any other questions.<span><font color="#888888"><br></font></span></div></div></blockquote><div><br></div></span><div>Thanks for your patient and informative explanation and suggestions, now I have better understanding of this assignment's target, and I'll draft a project proposal shortly. And hope to hear from you soon.<br><br></div><div>All the best,<br></div><div><br></div><div>Michelle<br></div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><span><font color="#888888"><br></font></span></div><span><font color="#888888">Jo<br><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><br></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></font></span></div><div class="gmail_extra"><br><div class="gmail_quote"><div><div>2015-03-19 13:04 GMT+01:00 Michelle W.X.Yang <span dir="ltr"><<a href="mailto:wx.michelle.yang@gmail.com" target="_blank">wx.michelle.yang@gmail.com</a>></span>:<br></div></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div><div dir="ltr">Hey,<div><br></div><div>I'm a CS PhD student from the University of Hong Kong, and my research is related to information visualization. I'm quite interested in learning more information about the project: JOSM Plugin to work with Mapillary images.</div><div><br></div><div>In my understanding, In this project, we need to develop a plugin for the OpenStreetMap which allows user to navigate as well as edit Mapillary images. In order to implement these functions, on the user interface of OSM, we need to develop a Mapillary layer and a Mapilary Window Pane, and we also need to develop some background components, e.g. cache. Even though I can get the main ideas of this project from the wiki page of Mapillary plugin for JSOM by referencing the website of Open Street Map and Mapillary, some details of the requirements as well as the scope of this project are needed to be clarified. I hope we can discuss about this project as soon as possible.</div><div><br></div><div>I have 3+ years of hand-in java experience, and since my research is related to information visualization, I'm quite familiar with developing user interfaces, image processing techniques and some graphics algorithms, I'm also very familiar with cache, e.g. LRU cache. Therefore, based on my background I believe I will be a good candidate for this project.</div><div><br></div><div>Sincerely,</div><div>Michelle.</div></div>
<br></div></div><span>_______________________________________________<br>
dev mailing list<br>
<a href="mailto:dev@openstreetmap.org" target="_blank">dev@openstreetmap.org</a><br>
<a href="https://lists.openstreetmap.org/listinfo/dev" target="_blank">https://lists.openstreetmap.org/listinfo/dev</a><br>
<br></span></blockquote></div><br></div>
</blockquote></span></div><br></div></div>
</blockquote></div><br></div>