基于空间分布特性和局部复杂度的显著目标检测算法

目前的视觉注意显著区域检测算法,主要依赖像素间的相互对比,缺乏从全局角度对显著目标的分析理解。依据显著目标是显眼、紧凑和完整的思路,提出一种基于目标空间分布特性和局部复杂度的无监督视觉注意显著目标自动提取算法。首先根据局部区域与其多个尺度邻域的对比,得到亮度显著图;然后利用颜色信息的显眼性、空间分布和区域一致性得到颜色显著图;同时通过对方向的空间分布和局部复杂度进行多尺度分析得到方向显著图;最后通过显著值的空间分布和面积增强因子相结合的融合策略得到输入图像的显著图,根据显著区域确定感兴趣区域位置,在基础上完成目标检测。将此方法应用于各类具有不同特点的彩色图像进行仿真实验,得到较好的检测结果,表明该算法是切实可行。

Salient object detection algorithm based on space distribution and local complexity

The current existing salient region detection algorithm mainly focus on the inter-pixel contrast and lack of the analysis and comprehension of salient object features from the global perspective.According to the thought that a salient object in a image is often compact and complete,a unsupervised salient object extraction algorithm based on space distribution and local complexity is proposed in this paper.First of all,the intensity saliency map is obtained by computing the contrast of local area and its neighborhood region,and then the color saliency map is computed using conspicuous,space distribution and regional homogeneity of color information.Meanwhile,the orientation saliency map is obtained by multi-scale analysis of the spatial distribution and regional complexity of the orientation.Finally,a integrated saliency map of the input image is composed through fusion strategy by combining the spatial distribution of salient value with enhancement factor of area,and the object of intere st is located and extracted according to the saliency map.The proposed method is applied to all kinds of color images,and the better test results show that the algorithm is feasible and valuable.