百分位形态变换在红外图像边缘检测中的应用

针对红外图像对比度低、边缘模糊、噪声大、空间相关性强的特点，提出了一种基于图像形态学的红外图像边缘检测方法。采用并行复合顺序形态滤波算法对图像边缘进行增强和锐化并抑制高频噪声；根据百分位形态滤波相关概念及性质构造的边缘检测算子，进一步滤除红外图像中的噪声，提取图像边缘。实验结果表明，本文提出的算法可以有效地克服红外图像的缺陷，保持图像边缘细节，优于传统边缘检测器。

Infrared Image Edge Detection Based on Percentile Morphological Transformation

A new method of infrared image edge detection based on image morphology was put forward aiming at the shortages of infrared image such as low contrast, blurred edges, serious noise and strongly spatial correlation. A parallel multiplex order morphological transformation-based algorithm was used for infrared image enhancement, edge sharpening and filtering. To extract edge and restrain noise of infrared image efficiently, morphology operators were constructed on the basis of the concept and the properties correlating to percentile morphology filtering. The experiment results show that the method presented is more effective than classical edge detection operators for overcoming the defects of infrared image and retaining details of image edge.