何贵, 钟羽中, 李振阳, 吴克江, 佃松宜. 一种双曝光度图像融合算法[J]. 微电子学与计算机, 2021, 38(3): 21-26, 32.
引用本文: 何贵, 钟羽中, 李振阳, 吴克江, 佃松宜. 一种双曝光度图像融合算法[J]. 微电子学与计算机, 2021, 38(3): 21-26, 32.
HE Gui, ZHONG Yu-zhong, LI Zhen-yang, WU Ke-jiang, DIAN Song-yi. Research on dual-exposure image fusion algorithm[J]. Microelectronics & Computer, 2021, 38(3): 21-26, 32.
Citation: HE Gui, ZHONG Yu-zhong, LI Zhen-yang, WU Ke-jiang, DIAN Song-yi. Research on dual-exposure image fusion algorithm[J]. Microelectronics & Computer, 2021, 38(3): 21-26, 32.

一种双曝光度图像融合算法

Research on dual-exposure image fusion algorithm

  • 摘要: 为克服传统图像增强算法在光照强度不均的密闭腔体内对图像增强不足和自然度保留能力弱的问题,本文提出了一种双曝光度图像融合的算法.首先通过图像熵最大化准则估计最佳曝光度,然后利用相机响应模型合成最佳曝光度图像,并基于自适应的权重评估方法融合弱光照图像和最佳曝光度图像,最后利用改进的BM3D (Block-matching And 3D Filtering) 方法对融合图像去噪.所提方法在公开数据集和GIS设备的密闭腔体内进行了验证.实验结果表明,本文方法能够增加图像熵,减少图像噪声,提高低光照部分的亮度,保护图像自然度.

     

    Abstract: To overcome the problem of insufficient enhancement and weak naturalness preservation ability of the image in a closed cavity with uneven illumination intensity, a double exposure image fusion algorithm is proposed in this paper. Firstly, the optimal exposure is estimated by the image entropy maximization criterion. Secondly, the camera response model is used to synthesize the optimal exposure image. Thirdly, the low-light image and the optimal exposure image are fused based on the adaptive weight evaluation method. Finally, the improved BM3D (block-matching and 3D filtering) method is used to denoise the fusion image. The proposed method is validated in a closed chamber of open data set and GIS equipment. Experimental results show that the proposed method could increase the entropy of the image, reduce the noise of the image, improve the brightness of the low-light part, and protect image naturalness.

     

/

返回文章
返回