摘要
背景不均匀对目标的检测造成了一定的困难,因此提出了一种自适用背景校正方法,通过改进Gamma校正函数,实现了对不同灰度值的不同程度的修正,使得背景校正更加均匀合理。并且能够根据图像灰度的分布,自适用地确定相关参数。仿真结果表明,该方法效果优于传统的直方图均衡化法、同态滤波法和以往改进的Gamma校正法。经过进一步改进将该方法应用于系留缆绳的损伤检测中,通过实验证明,经过背景校正的图像,利用迭代选择阈值法也可以获得理想的图像分割效果。
For the difficulty of target detection in uneven background, an adaptive background correction method is proposed. The method can correct different pixel values to various degrees by modified Gamma correcting functions as a result the background is more even and more appropriate. Furthermore, the method can determine correlative parameters according to the distribution of image＇s gray-level. The simulation re- sults demonstrate that our method is better than traditional histogram equalization method, homomorphic fil- tering method and modified Gamma correction method before. By farther modification, the method is used to detect mooring cable＇s defect. The experimental results show that the iterative threshold method can also provide good segmented image after background correction.
出处
《雷达科学与技术》
2012年第6期645-649,共5页
Radar Science and Technology
基金
总装维改项目
关键词
背景校正
GAMMA校正
系留缆绳
损伤检测
background corrections Gamma corrections mooring cables defect detection
