【摘要】 讨论了采用无监督的模糊竞争学习算法,并结合自组织竞争网络构成的一种新型模糊聚类神经网络模型,提出了一种基于该网络模型的镜头突变检测算法。该算法通过对线性特征空间进行由粗到细的两步模糊聚类实现镜头突变的检测。实验结果表明该算法是可行和有效的。更多还原

【Abstract】 A novel model of fuzzy clustering neural network is discussed, which synthesizes unsupervised fuzzy competitive learningalgorithm and self-organizing competitive network. Based on this model, an algorithm of video shot boundary detection is presented.This algorithm performs the detection by a two-stage clustering on the linear feature space. The experimental results obtained demonstratethat the algorithm is feasible and efficient.更多还原