一种基于视觉注意模型的可移动目标检测方法

提出了一种基于视觉显著性的目标检测算法，用于对位于地面上的车辆、飞 机等地面可移动目标进行检测和定位。针对地面可移动目标在场景中较小的特点，设计了一种用于对 目标检测进行引导的基于视觉注意机制的目标显著模型。首先，提取图像目标的颜色特征、强 度特征和方向Gabor特征，并将其结合起来用于计算显著图。然后以超像素为单位对 显著值进行计算，并结合人眼视觉敏感度对不同距离的超像素之间的差异进行加权处理。实 验结果表明，本文算法可以有效地检测并定位出复杂背景中的地面可移动目标。

A Ground Mobile Target Detection Method Based on Visual Attention Model

A target detection method for detection and location of mobile targets such as vehicles and airplanes on the ground based on a visual attention model is proposed. According to the factor that the mobile targets in natural scenes are small, a target saliency model for guiding target detection based on visual attention mechanism is designed. First, the features such as color, identity and orientation in a target image are extracted and are combined with each other to calculate the saliency map. Then, the saliency value is calculated in the unit of superpixel and a weight is provided for the difference among the superpixels at different distances by combining with the human visual sensitivity. The experimental results show that the algorithm can detect and locate mobile targets against complicated background on the ground effectively.