View Issue Details

IDProjectCategoryView StatusLast Update
0024179Open CASCADEOCCT:Modeling Algorithmspublic2016-10-28 22:00
ReporterifvAssigned Tobugmaster  
PrioritynormalSeverityfeature 
Status closedResolutionno change required 
PlatformAOSL 
Product VersionUnscheduled 
Summary0024179: Development of special method for intersection plane and parametric surface
DescriptionTask is performed according to OCCT 2013 development plan.
TagsNo tags attached.
Test case numberNot needed

Activities

kgv

2013-12-04 08:32

developer   ~0027062

Dear ifv,

this issue is registered as BUG.
Could you please extend description by relevant information (the bug description / steps to reproduce etc.)?

ifv

2013-12-04 09:57

developer   ~0027063

Dear kgv,
really it is not bug, but some kinds of imrovement, so for now it has no bug descriptions and step to reproduce

nbv

2016-10-24 11:44

developer   ~0059004

Dear Mikhail,

I think we should close this issue because advantages of new method over existing are insignificant. Namely:

1) Intersection line is planed to be represented by Hermite polynomial of the third degree (maximum). In our test base there are cases, where intersection line is represented by polynomial of the 10th degree and above. It means that (in general case) we will obtain here the intersection curve, which contains greater number of poles/knots. I.e. it does not lead to the intersection result improve. Nevertheless, we cannot use high order (degree is greater than 3) Hermite polynomials by the reason of small source data (for 3rd degree, it is required to know values and tangent vectors on the ends of the polynomial; for greater degree, additional conditions are required and cannot be obtained by simple methods).

2) Appearance of 3rd-degree Hermite polynomial depends strongly from lengths of tangent-vectors in the interpolation points. Existing analytical methods allow determining only directions of tangent-vectors (and accurate within sign). Use of optimization method is required in order to compute precise length values. However, solving of optimization problem requires some time. Attempts to estimate lengths (with Lagrange interpolation does not lead to significant improvements).

3) New algorithm will be oriented on cases where one of intersected surfaces is planar. Therefore, we will obtain a part of the code, which will work in private cases only (instead of existing universal intersector).

According to above said, we will not obtain significant improvements in simplification of intersection curve(s) and in speeding up.

msv

2016-10-24 12:03

developer   ~0059008

Dear bugmaster, please close it.

git

2016-10-28 22:00

administrator   ~0059612

Branch CR24179 has been deleted by kgv.

SHA-1: 8e31653a8358eaa2a3ae9c6688e7f25808b14943

Issue History

Date Modified Username Field Change
2013-09-16 12:18 ifv New Issue
2013-09-16 12:18 ifv Assigned To => nbv
2013-12-04 08:32 kgv Note Added: 0027062
2013-12-04 09:57 ifv Note Added: 0027063
2013-12-04 09:58 kgv Severity minor => feature
2013-12-04 09:58 kgv Reproducibility have not tried => N/A
2016-10-24 11:44 nbv Note Added: 0059004
2016-10-24 11:44 nbv Assigned To nbv => msv
2016-10-24 11:44 nbv Status new => feedback
2016-10-24 12:03 msv Note Added: 0059008
2016-10-24 12:03 msv Assigned To msv => bugmaster
2016-10-25 18:16 apn Test case number => Not needed
2016-10-25 18:16 apn Status feedback => closed
2016-10-25 18:16 apn Resolution open => no change required
2016-10-25 18:16 apn Target Version Unscheduled =>
2016-10-28 22:00 git Note Added: 0059612