Hyperpixels: Flexible 4D Over-segmentation for Dense and Sparse Light Fields

Maryam Hamad^1,2, Caroline Conti^1,2, Paulo Nunes^1,2 Luís Ducla Soares^1,2
1Instituto Universitário de Lisboa (ISCTE-IUL), ²Instituto de Telecomunicações


Published in IEEE Transactions on Image Processing: Link to our paper

The main steps of the proposed 4D LF over-segmentation method. Given a 4D LF and the corresponding disparity maps for all views, initial centroids characterized by distinct features are assigned in the reference view/views. Next, hyperpixels are generated by iteratively applying 4D K-means clustering, including pixel labeling, centroids adjustment and clustering weights adaptation, until convergence is reached.

The generated 4D light field dataset can be found in our group website:

Link to our IT-4DLF dataset

If you use this dataset in your work, please cite our paper:

M. Hamad, C. Conti, P. Nunes and L. D. Soares, "Hyperpixels: Flexible 4D Over-segmentation for Dense and Sparse Light Fields," in IEEE Transactions on Image Processing, doi: 10.1109/TIP.2023.3290523.

The dynamic results videos can be found in this repository:

Link to the dynamic results