利用视觉显著性的图像分割方法

Image segmentation based on visual saliency

提出一种利用视觉显著性对图像进行分割的方法。首先提取图像的底层视觉特征,从局部显著性、全局显著性和稀少性3个方面计算各特征图像中各像素的视觉显著性,得到各特征显著图;对各特征显著图进行综合,生成最终的综合显著图。然后对综合显著图进行阈值分割,得到二值图像,将二值图像与原始图像叠加,将前景和背景分离,得到图像分割结果。在多幅自然图像上进行实验验证,并给出相应的实验结果和分析。实验结果表明,该方法正确有效,具有和人类视觉特性相符合的分割效果。

An approach for image segmentation based on visual saliency is proposed in this paper. First low-level visual features of the image are extracted. Local saliency, global saliency and rarity saliency are computed for each feature map to get the feature conspicuity maps. Then these conspicuity maps are integrated to generate the saliency map. The saliency map is segmented using a threshold and a binary mask map is obtained, Finally the foreground and background of the original image are separated by adding the binary map to the original image. The proposed model has been tested on many natural images. Experimental results show that the proposed approach is valid and the segmentation results are consistent with human visual system.