改进的Retinex低照度图像增强算法

IMPROVED RETINEX LOW ILLUMINANCE IMAGE ENHANCEMENT ALGORITHM

  • 摘要: 针对低照度图像视觉效果差、图像不清晰、细节不能充分显现等问题,提出基于引导滤波的改进Retinex算法,使用引导滤波对低照度图像进行处理,对亮度分量进行自适应增强,再根据子图像特性进行加权融合。不仅能够保持图像边缘部分的细节,而且在增强图像暗区亮度的同时抑制局部高亮区的“光晕”。通过大量测试,该算法的改进不仅主观视觉效果上很明显,在峰值信噪比(PSNR)和结构相似性(SSIM)上也有较大提升。

     

    Abstract: In order to solve the problems of poor visual effect, unclear image and insufficient detail display in low-illuminance image, an improved Retinex algorithm based on guided filtering is proposed. Guided filtering was used to process low-illuminance image, and brightness component was enhanced adaptively, and weighted fusion was performed according to the characteristics of sub-image. It could not only preserve the details of the edge part of the image, but also suppress the "halo" of the local highlight area while enhancing the brightness of the dark area of the image. Through a lot of tests, the improvement of the algorithm is obvious not only in the subjective visual effect, but also in the peak signal to noise ratio (PSNR) and structural similarity (SSIM).

     

/

返回文章
返回