A multi-camera view synthesis known as Multi-Level View Synthesis (MLVS) approach based on Depth Image Layer Separation (DILS) algorithm for 3D vision and free-viewpoint video application is presented. The main aim of this technique is to create multi-camera view system through reduced number of actual image acquisition cameras, whilst maintaining the quality of the virtual view synthesis images. DILS separate the disparity depth map into several layers based on location and depth. Based on experimental results, the multi-camera of interview image synthesis provides good results compared with the conventional inter-view interpolation method. The proposed technique is shown to provide additional views with fewer cameras and lower complexity compared to conventional high volume camera configurations for free-viewpoint acquisition and light field imaging applications. Therefore, the multi-camera view synthesis can reduce the cost required.
- multi-camera view synthesis
- depth image layer separation
- multi-level view synthesis
Abd Manap, N., Mat Ibrahim, M., Soraghan, J., & Petropoulakis, L. (2014). Multi-camera view synthesis based on depth image layer separation (DILS) algorithm. Advanced Science Letters, 20(10-12), 1837-1841. https://doi.org/10.1166/asl.2014.5644