Abstract. The problems of low-contrast, blurred images, and low-edge details are encountered in infrared images. These undesirable features affect infrared imaging applications in military, science, medicine, and other fields. Therefore, a method of edge detection based on heat conduction matrix (HCM) has been proposed to show or sharpen the edges of objects in infrared images. First, the heat conduction formula of solid and still fluids was applied on the infrared image, and a feature matrix called HCM was obtained. Second, the HCM feature matrix was passed through a threshold to obtain distinct and noise-free edge images. The results of the numerical and visual applications were compared with the Sobel, Canny, and Laplacian of Gaussian methods. The results of our study showed that the proposed method is powerful in detecting edges in infrared images. The most important feature of the method is its high emphasis on important edges as well as thin and distinct edges.

中文翻译：

---

基于热传导矩阵的红外图像边缘检测方法

摘要。红外图像存在对比度低、图像模糊、边缘细节低等问题。这些不良特征会影响红外成像在军事、科学、医学和其他领域的应用。因此，人们提出了一种基于热传导矩阵（HCM）的边缘检测方法来显示或锐化红外图像中物体的边缘。首先，将固体和静止流体的热传导公式应用于红外图像，得到称为HCM的特征矩阵。其次，HCM 特征矩阵通过一个阈值以获得清晰且无噪声的边缘图像。数值和视觉应用的结果与高斯方法的 Sobel、Canny 和 Laplacian 进行了比较。我们的研究结果表明，所提出的方法在检测红外图像中的边缘方面是强大的。该方法最重要的特点是高度重视重要边缘以及细而清晰的边缘。