一种结合PE的高动态范围红外图像压缩及细节增强算法

针对在提升高动态范围红外图像中潜在或弱小目标细节的同时,还需兼顾噪声抑制、对比度增强的问题,提出了一种基于引导滤波图像分层的动态范围及细节增强算法.对背景层采用平台直方图均衡算法进行压缩,对细节层先采用中值滤波进行去噪,再采用非线性映射对细节中潜在的弱小目标细节进行增强,最后按照一定权重合并得到细节增强后的图像.综合主、客观实验结果,相对于映射类、直方图均衡、双边滤波分层增强等算法,该算法能够在动态范围压缩的过程中提高红外图像目标场景的对比度,突显其纹理特征,取得良好的细节增强效果.

Dynamic Range Compression and Detail Enhancement Algorithm Combined with PE for High Dynamic Range Infrared Images

To address the problem of noise suppression and contrast enhancement in enhancing the details of potential or small targets in infrared images with a high dynamic range,a dynamic range and detail enhancement algorithm based on a guided filter image is proposed.The base layer is compressed by a plateau-histogram equalization algorithm.The detail layer is denoised by a median filter first,and then the potential and small target details are enhanced by nonlinear mapping.Finally,the image is integrated according to a certain weight to obtain the enhanced details.Based on the subjective and objective experimental results,compared with algorithms based on mapping,histogram equalization,and a bilateral filter,this algorithm can improve the contrast of the target scene of the infrared image in the process of dynamic range reduction.The proposed algorithm can highlight the texture features and obtain a good detail enhancement effect.