Open skbizport opened 2 years ago
Hi,
The contact for questions and licensing is me, @.***
The functionality in GoTools for approximate implicitization address local approximation within intersection algorithms for separating to near intersecting curved surface that cannot be separated using convex hulls.
An approximation of the distance is the value divided by the length of the gradient.
To compute a signed distance of a CAD model, our approach would be to have an implicit 3-variate spline space based on LR B-splines/tensor product B-splines. This is one of the possible direction for a master thesis we expect to start soon.
With respect to the program code I cannot answer as this has been developed by my colleagues. I bbc this answer to them, but summer holiday has started, so I expect that they will first follow up in August.
Tor
Last ned Outlook for iOShttps://aka.ms/o0ukef
Fra: skbizport @.> Sendt: onsdag, juli 6, 2022 6:07 pm Til: SINTEF-Geometry/GoTools @.> Kopi: Subscribed @.***> Emne: [SINTEF-Geometry/GoTools] Few questions on implicitization (Issue #330)
Hi
I have few questions on approximate implicitization used in Go-Tools.
Regards SK
— Reply to this email directly, view it on GitHubhttps://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FSINTEF-Geometry%2FGoTools%2Fissues%2F330&data=05%7C01%7Ctor.dokken%40sintef.no%7Ccde0462917b446e9b82508da5f699b8b%7Ce1f00f39604145b0b309e0210d8b32af%7C1%7C0%7C637927204432951753%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=%2BtKapFl1eLBiRSuWPZNDQ8FZatXHs1rl31KJ5wNRI7s%3D&reserved=0, or unsubscribehttps://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FABBHBM2ZITYY6CF7T5CDTM3VSWVLRANCNFSM522JRJTQ&data=05%7C01%7Ctor.dokken%40sintef.no%7Ccde0462917b446e9b82508da5f699b8b%7Ce1f00f39604145b0b309e0210d8b32af%7C1%7C0%7C637927204432951753%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=DCQWKhh%2BwdJv2bSLzy97%2FYQdnNZcPGU9F9PavzI5ml4%3D&reserved=0. You are receiving this because you are subscribed to this thread.Message ID: @.***>
Appreciate these details. Would be quite interesting to see an implementation based on tri-variate splines.
Using the current library, it is not entirely clear how to compute the gradient for BernsteinTetrahedralPoly. Any pointers or example will be useful.
Also, is there any upper bound on error in the computed distance? Does it behave similar to euclidean distance?
Thx
Regards SK
Hi,
The master student we planned for working on this direction decided on another master thesis. However, I review a paper this summer (cannot give you details as it is not published) where gradient control was added to approximate implication in an elegant way. I believe that this idea can be used to better create distance fields using tri-variate tensor product B-splines or Locally Refined B-spliness. I think it is simpler if we communicate by ordinary e-mail, as what you are looking for is an extension of the functionality of GoTools. @.**@.>). I would also good to know who I am communicating with as we now enter a discussion of new research ideas and a not trivial extension of GoTools.
I am not the person implementing in GoTools, and the approximate implication functionality currently available targets the needs deep inside the recursive surface-surface intersection algorithm for deciding that two surface pieces don't intersect.
Tor
From: skbizport @.> Sent: onsdag 31. august 2022 02.56 To: SINTEF-Geometry/GoTools @.> Cc: Tor Dokken @.>; Comment @.> Subject: Re: [SINTEF-Geometry/GoTools] Few questions on implicitization (Issue #330)
Appreciate these details. Would be quite interesting to see an implementation based on tri-variate splines.
Using the current library, it is not entirely clear how to compute the gradient for BernsteinTetrahedralPoly.
Any pointers or example will be useful. Also, is there any upper bound on error in the computed distance? Does it behave similar to euclidean distance?
Thx
Regards SK
- Reply to this email directly, view it on GitHubhttps://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FSINTEF-Geometry%2FGoTools%2Fissues%2F330%23issuecomment-1232325608&data=05%7C01%7Ctor.dokken%40sintef.no%7C3b0e9e1ed57c48f5681408da8aeb9890%7Ce1f00f39604145b0b309e0210d8b32af%7C1%7C0%7C637975041724320057%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=VKauYls%2FTu8URyGvu6C8sHZjlcNQqmdBZXvegSb5pQQ%3D&reserved=0, or unsubscribehttps://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FABBHBM6TULQR22Q2LH34LA3V32USTANCNFSM522JRJTQ&data=05%7C01%7Ctor.dokken%40sintef.no%7C3b0e9e1ed57c48f5681408da8aeb9890%7Ce1f00f39604145b0b309e0210d8b32af%7C1%7C0%7C637975041724320057%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=jf2JR6kBHTqvUeCX6yuAIdXUjB%2FY4pgZA1Huu1ohD%2Fc%3D&reserved=0. You are receiving this because you commented.Message ID: @.**@.>>
Hi
I have few questions on approximate implicitization used in Go-Tools.
Regards SK