一种基于种群熵的自适应粒子群算法

在传统的含辅助信息水印检测方案中,通常的做法是从接收到的水印作品中减去原始作品,得到带噪声的水印模型.因此,水印不能完全提取.该文提出了一种基于预测攻击检测水印信号的方案,通过一定的方法预测水印图像所受到的攻击,根据预测结果,对原始图像进行同样的攻击来参与检测.实验表明该检测方案克服了辅助信息的残留噪声,大大提高了水印信息的检出率,改善了整个水印系统的性能.并且,该方案不失一般性,经验证可以应用到其他变换域和各种各样的算法中,有较强的通用性和应用前景.

Adaptive Particle Swarm Optimization Algorithm Based on Population Entropy

Matching is considered as one of the crucial factors to ensure dynamic discovery and composition of grid services. Current matching methods are inadequate given their inability to abstract and classify the underlying data of grid services. This paper proposes a semantic grid services discovery arithmetic which exploits FMD in order to abstract and classify the underlying data of grid services. Experimental results show FMD has more precision ratio and recall ratio.