铜导线断裂痕迹形态的图像特征提取与分析

Extraction and Analysis of Texture of Side Surface for Fracture of Copper Wires

纹理是断口的重要特征,不同的断裂原因造成断口的纹理特征是不一样的,它提示了图像中亮度值空间变化的重要信息。图像宏观上表现出来的是二维特性,可以将其灰度变化看成是图像的第三维特征来进行研究,这个"第三维"的表现可以用图像的能量或者所含信息量来表示。常用的图像纹理特征提取的方法有:统计法、利用空间自相关函数作纹理测度、频谱法、联合概率矩阵法、纹理的句法结构分析法。本文主要从频谱法和灰度共生矩阵法两个方面对4种不同原因造成的铜导线断口图像做纹理特征的提取与分析。

Texture is an important feature of fractures, different reason of rupture has different texture features, which contain lots information of directional distribution of image illumination. Image has a two-dimension feature macroscopic, and can call the diversity of gray scale die third-dimension of an image. This third-dimension can be described by capacity of infonnation. The methods of extraction and analysis of texture mainly include the following： statistical approach, lag correlation function approach, frequency spectrum, joint probability matrix and syntactic structures of texture. This paper mainly uses the frequency spectrum and gray symbiotic matrix to extract and analyze the features of texture of the four images of fracture of copper wires.