Synergetic Object Recognition Based on Visual Attention Saliency Map
Jing Shao, Jun Gao, Jing Yang · 2006
To study the object recognition in complex scene, a synergetic object recognition algorithm based on visual attention saliency map is proposed in the paper. We utilize the feature of the object extracted by PCA as the prototype vector of the synergetic pattern recognition. The adjoint vector is calculated through the synergetic learning algorithm. Then, the salient locations of the scene image including learned objects are selected through the visual attention saliency map. At last, the object in the salient location is recognized through the synergetic pattern recognition. The validity of the algorithm is demonstrated by the experiments