针对绝缘子红外与可见光图像融合过程中存在绝缘子伞盘边缘信息模糊, 亮度低和对比度差等问题, 本文提出了基于联合稀疏和参数自适应选择指导滤波的图像融合方法。图像首先经过联合稀疏模型分解, 提取共有特征、红外图像特有特征和可见光图像特有特征, 并按照特有特征系数的活跃程度调整权重; 同时应用参数自适应选择指导滤波方法, 能够较好地保留绝缘子图像的边缘信息和细节信息。通过对比实验, 本文方法融合结果亮度高、边缘清晰且边缘强度大, 同时客观指标也较好。

Aiming at the problem that the edge information of the insulator is fuzzy, the brightness is low and the contrast is poor in the fusion of the infrared and visible images, it is proposed that a method of image fusion based on joint sparse and parameter adaptive selection guided filtering in this paper. Images are first decomposed by joint sparse model, and common features, the unique features of infrared image and visible light image are extracted. The weights are adjusted according to the active degree of the characteristic coefficient. At the same time, this paper uses a parameter adaptive selection guided filtering method, which can preserve the edge information and detail information of the insulator image well. Through the contrast experiments, this method has shown high brightness, clear edge and large edge, and objective indicators are also better.