融合深度信息的视觉注意计算模型

针对Itti模型在特征选择上的局限性，借鉴心理学中有关视觉注意的研究成果，提出一种融合深度信息的视觉注意计算模型。该模型在基于图像分割的自适应立体匹配基础上提取深度特征，与亮度、方向、颜色特征相结合，实现空间显著性度量，并采用侧抑机制和WTA机制得到注意焦点。实验结果表明，新模型能较好地反映空间立体视觉信息对注意的影响，使计算结果能更符合人类视觉。

Depth Information Fused Computational Model of Visual Attention

The Itti model has a limitation on choosing features. Inspired by the visual attention results in the psychology, this paper proposes a computational model of visual attention. Based on the segment-based stereo matching which using belief propagation and a self-adapting dissimilarity measure, the depth information is computed. Intensity, color, orientation and depth information are deployed and create the saliency map. The saliency map is deployed to the FOA through inhibition of return and WTA mechanism. Experimental results show the model effectively reflects the influence of stereo vision to attention, and the computational result is more consistent with the human vision.