飞秒激光烧蚀光斑图像的双边滤波增强及FCM分割

飞秒激光烧蚀单晶硅材料过程中会伴随衍生等离子体发光,由CCD相机采集等离子体的光辐射信号可获得激光烧蚀光斑图像。针对光斑图像核心烧蚀区域易受光晕、烧蚀热过渡区域等因素影响问题,提出了烧蚀光斑图像的双边滤波增强结合模糊C均值聚类(FCM)分割策略,将光斑不同区域分割出来。首先,采用主成分分析(PCA)方法对光斑图像进行主成分提取,增强图像并去除噪声影响。其次,在PCA提取的第一主成分图像基础上,用双边滤波对光斑图像进一步增强处理,突出了光斑核心目标区域的层次。最后,采用模糊C-均值聚类(FCM)将光斑图像的各个区域精确的分割出来。通过与Canny算子结合高斯滤波的光斑图像分割效果对比表明,本文提出的光斑图像分割方法,对光斑核心区域、光斑细节及光斑层次分割方面具有优势。

Bilateral filtering enhancement and FCM image segmentation for ablated spot by femtosecond laser

In the process of femtosecond laser ablation of monocrystalline silicon material,the light emission of the derived plasma is accompanied,and the laser ablation spot image can be obtained by collecting the light radiation signal of the plasma from the CCD camera. Aiming at the problem that the core ablation area of spot image is easily affected by halo,ablation heat transition area and other factors,a bilateral filtering enhancement combined with fuzzy C-means clustering (FCM) segmentation strategy is proposed to segment different areas of spot image. Firstly,principal component analysis (PCA) is used to extract the principal component from the spot image to enhance image and remove the influence of noise. Secondly,on the basis of the first principal component image extracted by PCA,the spot image was further enhanced by bilateral filtering,which highlights the level of the core target area of the spot. Finally,fuzzy c-mean clustering (FCM) was used to segment each area of the spot image accurately. Compared with Canny operator combined with gaussian filter,the result of spot image segmentation shows that the proposed spot image segmentation method has advantages in the core area,spot details and spot level segmentation.