<html><head><meta http-equiv="Content-Type" content="text/html charset=utf-8"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class="">Hi all,<div class=""><br class=""></div><div class="">This is my week 12 report for multi-view reconstruction project. As usual, you can find the web version at <a href="https://wiki.blender.org/index.php?title=User:Tianwei37/Report_&_Blog/Week_12&action=edit&redlink=1" class="">https://wiki.blender.org/index.php?title=User:Tianwei37/Report_%26_Blog/Week_12&action=edit&redlink=1</a>.</div><div class=""><br class=""></div><div class="">== What I did this week: ==</div><div class="">* Following the work of last week, I have made the witness camera shown correctly. Users can use the 'secondary clip open' button to achieve this, which is in turn in the ‘correspondence’ mode of MovieEditor.</div><div class=""><br class=""></div><div class="">* Did some code refactor work to de-duplicate witness clip drawing code. The drawing code use just the main clip drawing code, with a flag (bool) in RegionSpaceClip to control which clip pointer should be used (‘clip' pointer or ‘secondary_clip' pointer)</div><div class=""><br class=""></div><div class="">* R<span style="font-family: HelveticaNeue;" class="">evise correspondence solve button because we no longer iterate over all views.</span></div><div class=""><span style="font-family: HelveticaNeue;" class=""><br class=""></span></div><div class=""><span style="font-family: HelveticaNeue;" class="">* Complete save/load (</span><font face="HelveticaNeue" class="">writefile.c and readfile.c) code for region data and secondary clips. Do versioning code for RegionSpaceClip, which solves the problem that dopesheet and graph (which are also two kinds of MovieEditors) cannot be opened. </font></div><div class=""><br class=""></div><div class="">The basic functionality of this multi-view reconstruction should work now. With the following limitations need to be solved, which I have discussed with my mentor:</div><div class="">1. Multi-view solver should be combined with single-view solver, leaving users with only one option.</div><div class="">2. There is an assertion bug related to transform.c code.</div><div class="">3. The multi-view solver can only work with two views for now. But this should be easy to fix.</div><div class="">4. There is a known bug related to saving/loading of correspondences, which needs further investigations. </div><div class="">5. Need to get witness cameras into the Scene. Now we only have one camera in the scene.</div><div class=""><br class=""></div><div class="">== What I will do next week: ==</div><div class="">* Next week marks the end of GSoC 2016. Cheers to all students (including me) who have made it to the end, and thank all mentors, especially my mentor Sergey, for your great helps! I will first finish with the assessment work, documentations, etc. related to GSoC. </div><div class=""><br class=""></div><div class="">* As I mentioned above, there are still some issues left to be solved. I will fix them as much as can next week. I will also invite artists (maybe on <a href="http://blenderartists.org" class="">blenderartists.org</a>, any other suggestions?) to test my branch when the pipeline is stable. I have done that once when the basic functioning multi-view solver is done. Thanks for the helps from artists but at that time the UI part is not done. With the new correspondence UI, it will be fun to test this new function. I also plan to make a video showing the pipeline. </div><div class=""><br class=""></div><div class=""><br class=""></div><div class="">Thanks,</div><div class="">Tianwei</div></body></html>