The paper first discusses shortcomings of classical adjacent-frame difference. Sec ondly, based on the image energy and high order statistic(HOS) theory, background reconstruction constraints are setup. Under the help...The paper first discusses shortcomings of classical adjacent-frame difference. Sec ondly, based on the image energy and high order statistic(HOS) theory, background reconstruction constraints are setup. Under the help of block-processing technology, background is reconstructed quickly. Finally, background difference is used to detect motion regions instead of adjacent frame difference. The DSP based platform tests indicate the background can be recovered losslessly in about one second, and moving regions are not influenced by moving target speeds. The algorithm has important usage both in theory and applications.展开更多
针对视觉背景提取(visual background extractor,ViBe)算法在运动目标检测过程中容易受到噪声干扰的问题,将两帧差分法融入ViBe的前景检测阶段,提出一种融合两帧差分信息的改进ViBe算法(ViBe with two-frame differencing,ViBe-TD)。首...针对视觉背景提取(visual background extractor,ViBe)算法在运动目标检测过程中容易受到噪声干扰的问题,将两帧差分法融入ViBe的前景检测阶段,提出一种融合两帧差分信息的改进ViBe算法(ViBe with two-frame differencing,ViBe-TD)。首先,设计单阈值形ViBe(single-threshold form of ViBe,S-ViBe)检测,为信息融合做准备;其次,基于逻辑斯蒂(logistic)回归模型,实现像素点上两帧差分和S-ViBe检测信息的融合;最后,综合两类检测信息完成前景像素点的判定。实验结果表明,ViBe-TD算法在4种不同场景视频上的检测效果达到了0.932的平均精确率,0.785的平均召回率以及0.842的平均F 1值。与原算法相比,ViBe-TD算法的各项指标平均有0.158的提高,具有良好的检测效果。展开更多
文摘The paper first discusses shortcomings of classical adjacent-frame difference. Sec ondly, based on the image energy and high order statistic(HOS) theory, background reconstruction constraints are setup. Under the help of block-processing technology, background is reconstructed quickly. Finally, background difference is used to detect motion regions instead of adjacent frame difference. The DSP based platform tests indicate the background can be recovered losslessly in about one second, and moving regions are not influenced by moving target speeds. The algorithm has important usage both in theory and applications.
文摘运动目标传统检测方法只考虑图像的亮度或纹理等某一种特性,受特异值影响较大,对噪声比较敏感,鲁棒性也不够好,而且背景恢复精度不高。针对以上局限性,提出一种融合结构相似度(structural similarity,SSIM)全参考模型和鲁棒主成分分析(robust principal component analysis,RPCA)的运动目标检测方法。此方法综合考虑图像的亮度、对比度和结构三种特性,不采用传统的背景减除法,而是把图像像素点的结构相似度作为度量来实现运动对象与背景的分离。实验结果表明,此方法准确率可达0.95,且F度量较传统运动目标检测算法平均提升0.15,总体上比传统方法更具优势。
文摘针对视觉背景提取(visual background extractor,ViBe)算法在运动目标检测过程中容易受到噪声干扰的问题,将两帧差分法融入ViBe的前景检测阶段,提出一种融合两帧差分信息的改进ViBe算法(ViBe with two-frame differencing,ViBe-TD)。首先,设计单阈值形ViBe(single-threshold form of ViBe,S-ViBe)检测,为信息融合做准备;其次,基于逻辑斯蒂(logistic)回归模型,实现像素点上两帧差分和S-ViBe检测信息的融合;最后,综合两类检测信息完成前景像素点的判定。实验结果表明,ViBe-TD算法在4种不同场景视频上的检测效果达到了0.932的平均精确率,0.785的平均召回率以及0.842的平均F 1值。与原算法相比,ViBe-TD算法的各项指标平均有0.158的提高,具有良好的检测效果。