利用者:Apinzonf/Gsoc2014/progress

提供: wiki
移動先: 案内検索

GSOC 2014 Progress Reports

Week 1 Report

What I did this week

This week I configure my workstation, need to get and install visual studio 2013, it had problems with the libraries for visual studio 2012.

I am copying and adapting the laplacian deform modifier to start testing that will allow me to define the initial scalar field.

I wrote and I tested in MATLAB code that implements the proposed document code Harmonic Functions for Quadrilateral Remeshing of Arbitrary Manifolds

Next week

Define a way to display and store the initial scalar field.


Questions

I do not know how to create a sort of groups of vertices but with positive and negative weights. as the first part of the process is to establish minimum and maximum points. And display to the user the initial scalar field in the same way that the "vertex groups".


Week 2 Report

What I did this week

I apologize. This week was completing several projects at my university, and I could not advance the project.

Next week

Define a way to display and store the initial scalar field.

Questions

.

Week 3 Report

What I did this week

I wrote a design document

User:Apinzonf/Gsoc2014/design


The desing documnet contain the deliverables.

I implemented a tool as a modifier, and use a vertex groups to define feature points.

The tool actually only compute the initial scalar field U, based on Harmonic Functions.

Next week

Write a document and a web page to get feedback from artist about the possible ways to interact with the tool.

Show some preliminaries results in a new vertex groups and capture some images of these results.

Questions

Is possible to change look up table of vertex groups, for use other colors.


.

Week 4 Report

What I did this week

I wrote a document to ask about user interaction. The most important step in remesh process is the construction of initial scalar field. The user needs to define in an easy way the maximum o minimum points and the weights. User:Apinzonf/Gsoc2014/test_scalar_field

I wrote a thread in BlenderArtist.org to discuss about user interaction.

http://blenderartists.org/forum/showthread.php?339897-GSOC-2014-QuadRemesh-How-do-you-think-it-should-be-the-user-interaction

The modifier put the results in a new vertex group named QuadRemeshGroup. The results are normalized between 0.0 – 1.0 to visualize this in a correct way in a vertex group created.

Next week

Compute the gradient scalar fields.

Questions

.

Week 5 Report

What I did this week

Since last week, my computer is very slow, i formatted my hard drive but the problem persists. I got a new 750gb hard drive. I am installing everything again. I had forgotten how to configure Git on Windows, and the ssh access keys. But all problems were solved.

I advised my mentor's problems since Tuesday

I wrote initial structures to compute the gradient field.

I wrote a method to compute the gradient field.

Next week

Continue calculating gradients fields

Questions

.

Week 6 Report

What I did this week

Structures were added to store the two gradient fields.

It ended the method to calculate the two gradients fields.

The results are now shown with only two colors, so that the scalar field is represented by bands.

Next week

Construction integral lines defined by the orthogonal vector fields.


Questions

.

Week 7 Report

What I did this week

I did not work this week, for personal family reasons and my mentor is aware.

Next week

Construction integral lines defined by the orthogonal vector fields.


Questions

.

Week 8 Report

What I did this week

I add a structure to manage the flow lines.

The main structure is modified to save a set of flow lines.

It created a method to calculate a flow line, given the index of a vertex that will serve as the seed of the line.


Next week

Continue doing the construction integral lines defined by the orthogonal vector fields.


Questions

.

Week 9 Report

What I did this week

Method for choose face that serve to build a line gradient flow.

Method to project vector field on certain face and intersect this with adjacent edges, this method present problem because i misinterpreted the m3 matrix representation.

Fixed problem about computation of gradient field vector.

Warnings fixed, about non used variables.

I changed the structure for managed flow lines of a gradient field.

Next week

Continue doing the construction integral lines defined by the orthogonal vector fields.


Questions

.

Week 10 Report

What I did this week

I wrote a method to store a map of edges around vertex.

I wrote a Method to project a gradient field vector on a plane defined by face for define the direction of flow line over this face.

I wrote a Method to intersect a face with line projected from seed point on gradient direction.

I had to solve several problems due to numerical problems is difficult to locate the intersections, and the creation of flow lines on the surface of triangles.


I lost a lot of time doing testing desk, very simple triangles and vectors to ensure the functioning of the basic parts of the algorithm.


The article did not mention such as flow lines were constructed, reason why I had to think of everything from seeds to creating conditions for the algorithm stops do not run in infinite loops.

Next week

Continue doing the construction integral lines defined by the orthogonal vector fields.


Questions

.

Week 11 Report

What I did this week

I wrote a wiki page to explain and an analyze the possible configurations when compute the flow line based on gradient projected over faces, this analysis avoid unnecessary calculation of intersections as much as possible.


User:Apinzonf/Gsoc2014/flowline


Based on the above document I wrote a method nextPointFlowLineInverse that is robust and fast. This method is exhaustive for calculating the path of the flow line.


I wrote a method to compute a random point, P, uniformly from within triangle ABC, method given by Robert Osada, Thomas Funkhouser, Bernard Chazelle, and David Dobkin. 2002. Shape distributions. ACM Trans. Graph. 21, 4 (October 2002), 807-832.


Next week

Continue doing the construction integral lines defined by the orthogonal vector fields.


Questions

.

Week 12 Report

What I did this week

I wrote a method to compute sampling distance functions, on the paper is not clear the way that how new flow line based on this distance functions are generated, because the computation of functions h1 are based on every vertex with the use of curvature defined how the dot product between normal at this vertex and the gradient field, but the gradient field was computed for every triangular face not every vertex, and other problem is that not explain how to use this sampling distance functions h1 and h2 to compute distance between line flows, how do you use this values (h1 on every vertex) inside a face, because the flow lines have different locations that vertices with h1 or h2 values.

To resolve this problem, i am reading these papers

Jobard, B. & Lefer, W.Creating Evenly-Spaced Streamlines of Arbitrary Density. Visualization in Scientific Computing ’97, Springer Vienna, 1997, 43-55

Mebarki, A.; Alliez, P. & Devillers, O. Farthest point seeding for efficient placement of streamlines Visualization, 2005. VIS 05. IEEE, 2005, 479-486

Rusinkiewicz, S. Estimating Curvatures and Their Derivatives on Triangle Meshes Proceedings of the 3D Data Processing, Visualization, and Transmission, 2Nd International Symposium, IEEE Computer Society, 2004, 486-493


Next week

Form a net of polygons over the surface by tracing integral lines of the vector fields G1 and G2.

Questions

.

Week 13

What I did this week

I am working on organize the code, and put the last pieces to compute the new mesh.

The project has a lot of different topics, and had several pieces of method not well defined on the paper. But was a challenge for me, and i learn a lot with this Gsoc.

I am writing a document about all project, with the completed parts and missing steps to complete the process, progress on the project is 90% and not long to finish it.

User:Apinzonf/Gsoc2014/finalreport

Next week

Working in the last steps of the project.

Questions

.