关键词： B样条, 图像融合, 熵, 梯度

Abstract: Image fusion is widely used in target recognition,computer vision,medicine,remote sensing,image processing and many other fields.This paper proposes an image fusion method based on B-spline gradient.It first divides the source image into small blocks of 3×3 size,calculates separately the B-spline gradient for each sub-block,adds up the gradient modulus values,calculates the weight value of each sub-block gradient modulus value,multiplies each sub-block and respective weight value and obtains the fusion image sub-block.Simulation experiments and analysis show that this method is effective and has a better performance than traditional methods such as principal component analysis,wavelet transformation and the Laplacian pyramid.

Key words: B-spline, image fusion, entropy, gradient