Bayes methods; Gaussian processes; cameras; image classification; image segmentation; image sequences; mixture models; object detection; Bayesian classification; Gaussian mixture model; MAP-MRF framework; background classes; background region; classification process; false negative detections; false positive detections; foreground classes; foreground object segmentation; foreground segregation; foreground-background probabilistic model; global background; graph-cut regularization; moving camera sequences; pixel-wise color GMM; prior probability maps; region-based models; Cameras; Computational modeling; Image color analysis; Image segmentation; Motion segmentation; Object segmentation; Probabilistic logic; Object segmentation; SCGMM; moving camera segmentation; spatial prior probability maps;
机译:交通监控系统中基于运动对象反馈的背景建模与前景分割方法
机译:可变形的概率图:概率形状和基于外观的对象分割
机译:基于迟滞和双向帧间变化检测的基于运动的对象分割,以及移动摄像机的序列
机译:基于前景背景概率模型和现有概率图的移动相机序列的前景对象分割
机译:对运动相机观察到的包含运动对象的场景进行运动分割和密集重建。
机译:基于多帧同形像元约束的静止或运动相机无监督运动对象分割
机译:具有非均匀背景和前景的物体的概率密度 ud使用基于概率密度函数的数据项和非参数形状先验来分割不均匀的前景和背景强度对象