摘要
应用适用于现场暴露试样的三维腐蚀形貌图像采集系统对Q235钢在模拟大气环境中的腐蚀形貌图像进行了采集。所得图像统一转换为灰度图像,进行中值滤波处理。为了更好的观察图像的细节信息,根据熵标准应用小波包变换对预处理后的图像进行分解,选取最优子图像后并提取其能量值。结果表明,根据图像能量值的变化可以对腐蚀试样进行定性的判断,而进一步根据计算出的图像特征值可以定量判断大气腐蚀早期腐蚀程度。
The corrosion morphology of the Q235 steel in simulated atmospheric environment was obtained using the established image acquisition system. In order to gain the detailed information of the images, wavelet packet-based image analysis method was applied to decompose the improved images and energies of sub-images (according to the Shannon entropy standard to choose the optimal sub-images with high regularity) were extracted as characteristic information. The results showed that the corrosion extent of the Q235 steel specimens was qualitatively analyzed which correlated well with the change of the energies of sub-images. The relationship between the characteristic information and corrosion mass loss of the samples was evaluated by regression analysis. Using the equation given by the regression analysis, the corrosion degree of the Q235 steel specimens could be quantitatively analyzed.
出处
《中国腐蚀与防护学报》
CAS
CSCD
北大核心
2013年第3期211-215,共5页
Journal of Chinese Society For Corrosion and Protection
基金
国家自然科学基金重点项目(51131007)
国家科技支撑计划项目(2011BAF11B08)
天津市自然科学基金项目(10JCYBJC03200)资助
关键词
大气腐蚀
腐蚀形貌
小波包变换
图像分析
腐蚀预测
atmospheric corrosion
corrosion morphology
wavelet packet transformation
image analysis
corrosion prediction