This paper proposes a complete and efficient video in-painting system, which is the process of removing a portion of a video and filling in the missing part (hole) in a visually consistent manner.
The described system applies different strategies to handle static and dynamic portions of the hole. To in-paint the static portion, the system uses background replacement and image inpainting techniques. To in-paint moving objects in the hole, it utilizes background subtraction and object segmentation to extract a set of object templates and perform optimal object interpolation using dynamic programming. The authors evaluate the performance of their system based on a set of indoor surveillance sequences with different types of occlusions. (Publisher abstract provided)
Downloads
Similar Publications
- Deep Hyperspectral Image Fusion Network With Iterative Spatio-Spectral Regularization
- Multi-Modal Analysis of Body-Worn Camera Recordings: Evaluating Novel Methods for Measuring Police Implementation of Procedural Justice
- Look Twice as Much as You Say: Scene Graph Contrastive Learning for Self-Supervised Image Caption Generation