Week 19 (May. 4 - May. 10): Investigation on Similar Projects

Since we have another final project on web programming due this week, we spent some more time on discovering others’ work on tree generation using space colonization algorithm, in which we mostly focused on Baptiste’s work, since it includes detailed explanation and clear coding.

Throughout reading Baptiste’s work, we were inspired by several points.

1. The starting point of our coding could be the definition of the branch class.

2. The distribution of attraction points in the crown volume should be uniform. Baptiste previously used linear randomization in a spherical coordinate system for a sphere crown volume, which let the attraction points be more centered and finally led to a non-real tree generation.

Spherical coordinate system (left), attraction points generated by a linear randomization (right)

1. The problem of discontinuity between branch mesh mentioned in our project specification could be resolved by Baptiste’s method

A possible method for branch mesh generation

1. The leaf could be generated at the position of attraction points.

2. We should expect our project to be a gradually-improved procedure. Included parameters could be often adjusted to ensure better results.

Work distribution:

Ying: branch mesh generation

Yuwen: attraction point distribution

Ziqi: space colonization algorithm