Towards a Fully Automated Tissue Cell Segmentation Method
| Authors | |
|---|---|
| Year of publication | 2004 |
| Type | Article in Proceedings |
| Conference | Visualization, Imaging, and Image Processing |
| MU Faculty or unit | |
| Citation | |
| Web | http://www.actapress.com/PaperInfo.aspx?PaperID=18789&reason=500 |
| Field | Informatics |
| Keywords | Biomedical image segmentation; Adaptive thresholding; Gradient based thresholding; Region labeling |
| Description | In this paper, an automatic method for tissue cell localization is proposed. It is based on the principle of adaptive thresholding. While the general adaptive thresholding method requires setting of some parameters by user, a novel algorithm chooses suitable thresholds and image division on its own. The resulting binary image is then submitted to a region labeling procedure. Each individual region is inspected and tested for suitability for further processing. Finally, each region is marked by one marker. Feasibility of this approach has already been successfully tested on 3D colon tissue images. |
| Related projects: |