›› 2012, Vol. 25 ›› Issue (3): 1-.

• 论文 •    下一篇

一种鲁棒的低对比度图像预处理算法

张俊生,王明泉,侯慧玲,陈志强   

  1. (1.中北大学 仪器科学与动态测试教育部重点实验室,山西 太原 030051;2.太原工业学院 电子工程系,山西 太原 030008)
  • 出版日期:2012-03-15 发布日期:2012-03-21
  • 作者简介:张俊生(1981—),男,硕士研究生,助教。研究方向:射线无损检测,图像识别和机器视觉。
  • 基金资助:

    山西省青年科技研究基金资助项目(2009021019-2);教育部高校博士点专项科研基金资助项目(20100501JJ)。

A Robust Pretreatment Algorithm for Low Contrast Images

 ZHANG Jun-Sheng, WANG Ming-Quan, HOU Hui-Ling, CHEN Zhi-Qiang   

  1. (1.Ministry of Education Key Laboratory of Instrumentation Sciences and Dynamic Testing,North University of China,Taiyuan 030051,China;
    2.Department of Electronic Engineering,Taiyuan Institute of Technology,Taiyuan 030008,China)
  • Online:2012-03-15 Published:2012-03-21

摘要:

数字图像在采集过程中,由于照明或物体本身等原因,经常会出现目标区域对比度低、细节模糊的情况,需要经过进一步的预处理才能使用。灰度分段线性变换是一种较为有效的对比度拉升技术,它通过选择合理的分段点来实现图像增强。一种常见的分段点选择方法是以原始图像灰度值的最大和最小值为分段点,但它对整体动态范围较大的图像视觉效果改善有限。文中通过对图像直方图的分析,在其累积直方图上设定阈值来确定分段点是一种可行的方法,具有更好的鲁棒性。仿真实验结果表明,该算法能够有效地增强低对比度区域,为图像的进一步处理提供理想的结果。

关键词: 对比度;线性变换;分段点;直方图

Abstract:

Due to illumination or the object itself,images often become low contrast or blurred in the acquisition process and need further pretreatment before they are used.Grayscale subsection linear transformation is an effective technology for contrast enhancement by means of the rational choice of subsection points.The maximum and minimum values are always used for subsection points,but this method does not work when the original image has wide dynamic range of gray value.Through the analysis of the original image histogram,a robust method is proposed which selects subsection points by setting thresholds in the accumulation of the gray histogram.Experimental results show that the method can enhance the region of low contrast and provide satisfactory images for further processing.

Key words: contrast;linear transformation;subsection point;histogram

中图分类号: 

  • TP391.41