###
DOI:
电力大数据:2020,23(5):-
←前一篇   |   后一篇→
本文二维码信息
低照度液晶屏幕图像增强算法
张薇薇1, 王彦1, 张庆伟1, 付龙明2, 黄辉2
(1.贵州电网有限责任公司贵阳供电局;2.中恒博瑞数字电力科技有限公司)
Image enhancement algorithm of low illumination LCD screen
zhangweiwei1, wangyan1, zhangqingwei1, fulongming2, huanghui2
(1.Guiyang Power Supply Bureau of Guizhou Power Grid Co., Ltd;2.Beijing Join Bright Digital Power Technology Co.Ltd)
摘要
图/表
参考文献
相似文献
本文已被:浏览 481次   下载 3
投稿时间:2020-04-28    修订日期:2020-06-05
中文摘要: 为解决目前自然场景文本识别系统文字识别率较低的问题,从识别图像的质量角度分析,发现用相机拍摄的带有液晶屏幕的图像,易受到光照或屏幕本身显示问题影响,成像质量低,难以辨别屏幕内信息。传统的解决方法如使用retinex算法进行图像增强时,由于retinex算法本身的局限性,对于光照不均的图像,会出现光晕现象。针对此问题,本文提出基于同态滤波的去光晕retinex算法来进行光照的校正,首先使用改进的同态滤波算法对图像进行不同程度的滤波处理,校正光照,处理后图像再使用改进的retinex算法进行图像增强,则可以避免光晕现象。实验结果表明,本文方法可以有效提高低照度液晶屏幕图像的视觉效果,优于传统的retinex算法,无光晕现象产生,进而提高自然场景文本识别系统的识别准确率。
Abstract:the image with LCD screen taken by camera is easy to be affected by the illumination or the display problem of the screen itself. The image quality is low and it is difficult to distinguish the information in the screen. When using the traditional Retinex algorithm for image enhancement, there will be halo phenomenon due to uneven illumination. In order to solve this problem, this paper proposes to use homomorphic filtering to correct the illumination, and then use Retinex algorithm to enhance the image after the illumination correction, which can avoid the halo phenomenon. The experimental results show that this method can effectively improve the visual effect of low illumination LCD screen image, which is better than the traditional retinex algorithm
文章编号:     中图分类号:    文献标志码:
基金项目:
引用文本: