In the program of Master 2 (UPMC / Telecom ParisTech)
We have a couple of group or individual projects.
In this article, I introduce projects that i am currently working on.
- 3D Woven clothes rendering
- Recognition of Shapes for the Interpretation and image analysis
- Image segmentation using Mathematical morphology techniques
The aim of this project is to implement an algorithm for rendering fabrics in real time.
This project is divided into two parts :
- The first part consists in making a bibliographical study. I have to read a couple of paper about the subject, then choose one method that i will have to exprain in detail.
- The second part consists in implementing the method choosen in the first part.
Expected results :
At the end i will have to write a report (20pages) where i'll explain the whole work, and show the results.
This project is split into 2 sessions :
- Calculation of local descriptors and visual dictionary
- Extract discriminative local features of each region of the image.
- Develop a clustering algorithm (K-Means) in the space of descriptors to produce a visual dictionary then to effectively compare different images.
The ultimate goal of the two sessions is to put in place a system image classification by supervised learning.
Once learned the system, it will be able to assign one of 15 semantic categories in a given test image.
Image segmentation by watershed under constraints