金相图像均衡化和多尺度模板二值降噪方法研究

Equilibrium of Metallographic Image and the Research of Multiscale Template Binary Noise Reduction

金相图像是金属材料微观形貌的体现,对金相图像进行分析是研究金属材料的重要手段。针对金相图像噪声强、晶界弱、光照不均匀等特点,结合数字图像处理方法,提出了一种金相图像二值降噪方法。首先,采用一种预处理综合方案对图像进行均衡处理。然后采取自适应阈值的方法得到金相二值图像,但仍存在较多噪声。对此,提出一种基于多尺度模板的降噪方法,该方法能有效去除大部分噪声。结果表明,该方法能够得到较好的图像预处理结果和清晰且较为准确的二值化结果,这将有利于后续金相定量分析工作的开展。

The metallographic image was the embodiment of the metal surface morphology.The analysis of the metallographic image was an important means to study the metal material.Aiming at the characteristics of metallographic image noise,weak grain boundary and uneven illumination,combined with digital image processing method,a binary image denoising method for metallographic image was proposed.First of all,a pre-processing integrated scheme to equalize the image was carried out.The adaptive threshold method was then used to obtain the metallographic binary image,but there was still more noise.In this regard,a multi-scale template based noise reduction method was proposed,which can effectively remove most of the noise.The results show that the proposed method can obtain better image preprocessing results and clear and accurate binarization results,which will be beneficial to the subsequent metallographic quantitative analysis.