煤炭工程 ›› 2024, Vol. 56 ›› Issue (11): 182-187.doi: 10.11799/ce202411027

• 研究探讨 • 上一篇    下一篇

基于多透射率融合的煤矿井下图像清晰化研究

董俊亮,张彦董,崔颢   

  1. 1. 神华新街能源有限责任公司
    2.
  • 收稿日期:2023-12-05 修回日期:2023-12-22 出版日期:2023-11-20 发布日期:2025-01-03
  • 通讯作者: 董俊亮 E-mail:10034638@ceic.com

  • Received:2023-12-05 Revised:2023-12-22 Online:2023-11-20 Published:2025-01-03
  • Contact: Jun LiangDong E-mail:10034638@ceic.com

摘要: 针对煤矿井下环境昏暗,目标模糊不清等问题,提出一种基于多透射率融合的煤矿井下图像清晰化方法。首先结合图像去雾方法,对图像HSV通道进行最小滤波处理,并取最大值并转换为RGB值作为大气光,然后提出多透射率融合方法,通过求多尺度暗通道图像,获取与多个尺度暗通道相对应透射率,并设计融合方法将多个透射率进行融合,最后利用获得的大气光和透射率,根据大气散射模型,求得清晰化图像,实现煤矿井下图像清晰化。实验证明,本文方法能有效提升图像质量,在主观上能有效的改善图像的细节信息,提升图像整体亮度,客观上在信息熵、平均梯度、模糊系数及结构相似度等指标上均取得明显改善。

关键词: 图像清晰化, 多透射率, 大气光, 图像增强

Abstract: Aiming at the problems of dark environment and blurred target in coal mine, this paper presents a method of image sharpening in coal mine based on multi-transmittance fusion. First of all, combined with the image de-fogging method, the image HSV channel is processed with minimum filtering, and the maximum value is taken and converted to RGB value as atmospheric light. Then, the multi-transmittance fusion method is proposed. By obtaining the multi-scale dark channel image, the transmittance corresponding to the multi-scale dark channel is obtained, and the fusion method is designed to fuse multiple transmittance. Finally, using the atmospheric light and transmittance obtained, the clear image is obtained according to the atmospheric scattering model to realize the clear image in the coal mine. Experiments show that the proposed method can effectively improve the image quality, subjectively improve the details of the image, enhance the overall brightness of the image, and objectively improve the information entropy, average gradient, fuzzy coefficient and structural similarity.