The Digital Image and Video Processing (DIVP) Group was formed in 2006 at the University of Brasilia. Its members include faculty members from the Departments of Computer Science, Electrical, Software and Electronic Engineering, as well as graduate and undergraduate students.
The Group’s activities are centered around research and teaching in the areas of digital image and video processing, with the following major research areas.
Plenoptic point cloud
Point clouds are 3D representations of real-world objects, people and scenes, which can be conveyed in real time to remote locations, enabling free-viewpoint viewing and rich collaboration as if all parties were co-located. The Group has developed compression technology that is currently adopted by MPEG’s Geometry-based Point Cloud Compression (G-PCC) standard. Other research topics include:
Geometric partition of blocks for motion estimation and compensation
The Group started researching several aspects of image and video compression, such as:
- Encoding complexity
- Wyner-Ziv coding
- Scanned document compression
- Intra prediction
- Motion estimation
Furthermore, our work on non-rectangular block partition was later adopted by the AV1 open video coding standard.
JPEG compression history estimation for color images
Besides image and video compression, our Group has also developed a large array of image and video processing technologies, such as: