This paper discusses the approaches for automatical searching of control points in the NOAA AVHRR image on the basis of data rearrangement in the form of latitude and longitude grid. The vegetation index transformatio...This paper discusses the approaches for automatical searching of control points in the NOAA AVHRR image on the basis of data rearrangement in the form of latitude and longitude grid. The vegetation index transformation and multi-level matching strategies have been proven effective and successful as the experiments show while the control point database is established.展开更多
Mean Shift跟踪算法在目标尺度变化大和被遮挡时存在较大的缺陷。针对这一问题,提出了一种基于多级正方形匹配的自适应带宽选择和分块抗遮挡的目标跟踪算法。该算法采用目标中心点的离散程度和增量试探法计算出可能的变化尺度,然后采用...Mean Shift跟踪算法在目标尺度变化大和被遮挡时存在较大的缺陷。针对这一问题,提出了一种基于多级正方形匹配的自适应带宽选择和分块抗遮挡的目标跟踪算法。该算法采用目标中心点的离散程度和增量试探法计算出可能的变化尺度,然后采用多级正方形匹配法预测目标的运动趋势,将巴氏系数最大者的尺度作为Mean Shift核函数新的带宽。同时,对前景目标进行分块,根据子块的遮挡程度自适应改变子块权重并按一定准则融合有效子块的跟踪结果。实验结果表明,该算法具有很好的鲁棒性。展开更多
基金Project supported by the National Oommission of Defense Science and Technotocjy(No.Y96-10)
文摘This paper discusses the approaches for automatical searching of control points in the NOAA AVHRR image on the basis of data rearrangement in the form of latitude and longitude grid. The vegetation index transformation and multi-level matching strategies have been proven effective and successful as the experiments show while the control point database is established.
