中文摘要：

红外图像包含物体的温度信息，但其存在对比性差、纹理弱等缺陷，限制了应用，目前基于融合的方法能有效改善红外图像的视觉效果，但局限于简单的直接融合，忽略了背景等因素所含噪声的影响及各部分细节信息。文章在这方面做了进一步的研究工作，改进了现有方法的融合规则，提出先将目标从背景中提取出来再以温度阈值及纹理特征为依据分层次分区域融合，从而在细节上极大地改进了目标的视觉效果，提高了效率。最后对融合效果进行了定量评价和比较。实验结果证明处理后的图像能够比原图像获得更丰富的视觉信息。

英文摘要：

Infrared image contained temperature information, but its application is greatly restriction due to inherent defect, such as bad contrast and weak veins etc. Though the present method of fusion can improve the visual effect of infrared image ,it is still confined to single and direct fusion which neglect the influence of background＇s yawp factor. The paper makes farther research on the respect, improve the existence fusion＇s regulation through propose a method that extraction the objective firstly then fuse the images layered which based on temperature threshold and texture characteristic. Thereby the specific of visual and effect are improved largely. Finally, evaluate the effect of the result through objective mathord. The result shows that our method can obtain more information than the original image.