计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (25): 165-167.

• 图形、图像、模式识别 • 上一篇    下一篇

侧扫声纳图像分割中种子区域获取方法

卞红雨,罗明愿,刘立昕   

  1. 哈尔滨工程大学 水声技术国家级重点实验室,哈尔滨 150001

  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-09-01 发布日期:2011-09-01

Method for getting seed region of side-scan sonar image segmentation

BIAN Hongyu,LUO Mingyuan,LIU Lixin   

  1. National Laboratory of Underwater Acoustic Technology,Harbin Engineering University,Harbin 150001,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-09-01 Published:2011-09-01

摘要: 研究了一种基于区域增长的侧扫声纳图像感兴趣区域提取方法。该方法利用边缘检测算子提取侧扫声纳图像的目标影阴区的种子区域,利用阈值分割方法提取目标亮区的种子区域,采用Haralick与Shapiro提出的相似准则进行区域增长,提取感兴趣区域。从处理结果看,该方法对某些类型的侧扫声纳图像的感兴趣区域提取有较好的效果,而且种子区域的引入大大缩短了处理时间。

关键词: 区域标记, 感兴趣区域, 侧扫声纳图像分割, 相似准则, 种子区域

Abstract: This paper presents a method for regions of interest extraction of side-scan sonar image based on region growing.Using edge detector,the seed regions of shadow region are obtained;and using threshold segmentation,the seed regions of high-light area are got.Adopting the Haralick similarity criterion,the regions of interest are extracted based on region growing.Results show that this method has a perfect performance for some types of side-scan sonar image.And the method is timesaving relative to Haralick region growing algorithm because of the application of seed regions.

Key words: region labeling, regions of interest, side-scan sonar image segmentation, similarity criterion, seed region