Image Segmentation of Soybean Based on H-Dome(PDF)
《大豆科学》[ISSN:1000-9841/CN:23-1227/S]
- Issue:
- 2013年06期
- Page:
- 821-824
- Research Field:
- Publishing date:
Info
- Title:
- Image Segmentation of Soybean Based on H-Dome
- Author(s):
- ?SUN Xiao-ting; CHEN Jiang-hong; CHEN Qing-zhou
- ?(School of Information Science and Engineering,Fujian University of Technology,Fuzhou 350108,China)
- Keywords:
- Soybean seed; Image segmentation; Watershed; H-Dome
- PACS:
- -
- DOI:
- 10.11861/j.issn.1000-9841.2013.06.0821
- Abstract:
- ?During quality inspection of soybean，the adhesion of soybean seeds in image brings a great difficulty to the posterior detection and analysis.Accordingly，a watershed algorithm based on H-Dome was put forward.After the local minimums were merged by the grayscale morphological reconstruction of H-Dome，the distance transform of reconstructed image was obtained.And then the watershed algorithm was used to segment the image.At last，the watershed line was extracted to obtain the final segmentation image.The segmentation ratio of proposed algorithm reached 96.6%.
Last Update: 2014-04-04