【摘要】 针对同一部位源于不同成像原理的医学图像融合问题,提出了一种基于脉冲发放皮层模型(SCM)的自适应医学图像融合方法.不同于传统的神经网络模型,SCM具有更直观的运行机制和更优良的图像处理效果,并且待定参数较少,极大地降低了SCM的计算复杂度.同时使用赋时矩阵解决了经典SCM中迭代次数难以确定的缺陷.实验结果表明本文方法是有效可行的.更多还原

【Abstract】 An adaptive method for medical images based on spiking cortical model(SCM)has been proposed to conduct the medical images fusion on the same body part from different imaging principles.Different from the existing traditional neural networks models,SCM owns more subjective function mechanism and much better performance on image processing.Moreover,fewer parameters required setting lead to the great decline of the computational complexity.The time matrix is utilized to deal with the drawback of the uncertain iteration numbers in the classic SCM.Experimental results verify the effectiveness and feasibility of the proposed method.更多还原