几种边缘检测算子在红外图像处理方面应用研究

1.西安科技大学 电气与控制工程学院,陕西 西安 710054; 2.西北工业大学 自动化学院,陕西 西安 710029

红外图像; 边缘检测; 梯度算子

Application of several edge detection operator in infrared images processing
YUE Gai-li1,WANG Dong2,YANG Rui2

(1.College of Electrical and ControlEngineering,Xi'an University of Science andTechnology,Xi'an 710054,China; 2.School of Automation,Northwestern Polytechnical University,Xi'an 710072,China)

infrared image; edge detection; grad operator

备注

为了从原始红外图像中提取目标轮廓信息,并克服图像对比度低、噪声干扰大等缺点,需要对红外图像进行边缘检测处理。文章分析Sobel,Canny等几种常用边缘检测算子原理并将其应用于红外图像的边缘检测,同时根据仿真结果比较了它们所代表的一阶梯度算子、二阶梯度算子以及复合检测算子在不同类型红外图像边缘检测中的表现,总结了不同算子检测效果的特点并分析其对应的理论依据。为今后红外图像处理中边缘检测算子的选择提供了参考。

In order to get the edge information from original infrared image,which is featured by lower contrast ratio and loud noise,edge detection is a necessary and important processing approach to improve the effect of image.In this paper,several kinds of common edge-detection operator like Sobel,Canny are applied to edge detection of infrared image,their outcomes are compared and analyzed to explain the difference on detecting edge by different grad operator.The advantages and disadvantages of the operators are also explained to provide a reference for futher work.