摘要
在对连通性要求比较高的图像进行分割时,区域生长方法有独特的优势。在图像复杂的情况下,希望计算机可以快速并且完整地自动选取种子点。利用二维OTSU对图像进行预分割并以此作为选取种子点的依据,可以有效解决使用区域生长分割图像时种子点选取的难题。结合对石化炼油炉炉管图像的分割,讨论使用此种方法的可行性和有效性。
For image segmentation in which connectivity requirements are relatively high, region growing method has unique advantages. In the complex situtaion, that the computer can fast, completely and au- tomatically select seed points is needed. Using two-dimensional OTSU pre-segmentation of the image can effectively bined the method is solve the problem of seed points selection by using region growing image segmentation. Com- image segmentation with petrochemical refining furnace tube, the feasibility and effectiveness of this discussed.
出处
《大气与环境光学学报》
CAS
2013年第6期448-453,共6页
Journal of Atmospheric and Environmental Optics
基金
科技部创新基金项目(07C26213400516)资助
