空地导弹在制导末段捕捉红外图像时,受成像器件的限制以及大气辐射影响,红外图像往往存在对比度差、边缘模糊、噪声大等特点。针对这些问题,文中将多尺度Retinex算法应用到空地导弹红外图像增强中,给出相应的实现方法和流程,并针对存在的不足加以改进。实验结果表明,改进后的算法在改善红外图像色调、对比度以及增强细节信息、去噪方面有较好的效果,使增强后的红外图像更能满足空地导弹末段的精确打击能力。

When an air-to-ground guided missile captures the infrared image at the terminal phase of guidance, the gained infrared images are bad in image contrast, blurry, large yawp because of the restriction of infrared imaging component and the radiation of atmosphere. Based on the problems, the multi-scale Retinex algorithm was applied to the air-to-ground guided missile's infrared image enhancement, and the corresponding method and flow were given. An improved algorithm was explored for the defects of the multi-scale Retinex. The result of the experiment shows that the improved algorithm has good effect on ameliorating infrared image hue, contrast, details and yawp removing, and which can enhance the precise hitting capability of the air-to-ground guided missile.