Ant colony optimization inspired saliency detection using compressed video information

A novel visual saliency detection algorithm using ant colony optimization and spatiotemporal information in compressed videos is proposed in this paper. Firstly, a graph is constructed for each frame in the video by dividing it into blocks and taking the block as nodes. We extract spatial and temporal features of each node directly from the compressed bitstreams to form the heuristic matrixes. Each heuristic matrix is used to steer the ants and the ants deposit pheromone on the graph. Then the pheromone is updated through attenuation and evaporation thus forming a spatial/temporal saliency map. Finally, an adaptive fusion method is used to merge the spatial and temporal saliency maps together. The proposed method has been extensively tested on several video databases with sequences in various scenes and experiment results show that it outperforms various state-of-the-art models in both quantitative evaluation scores and intuitive visual effects.