Cells were segmented using a custom-made image processing pipeline. The segmentation pipeline was implemented in order to distinguish cells from the background. The segmentation pipeline is composed of standard image-processing operations in the following order: 1, original image; 2, Sobel edge detection; 3, image dilation; 4, removal of objects close to image borders; 5, image erosion; 6, removal of small objects; 7, filling of gaps inside the cell; and 8, overlay of the final result on the original image.
Seven morphological features were extracted from each of the segmented cells. The feature space in which we performed statistical classification was therefore seven-dimensional (7D; one vector for each cell), with the following features: area, major and minor axis lengths, perimeter, eccentricity, extent, and number of fingers (Gorelick, PAMI, 2006). Statistical analysis was performed on the 7D feature vectors, using a tree-like classification method called the ’node harvest’ method, which was introduced by Meinshausen, Annals of Applied Statistics, 2010.
引用格式
Christof (2024). Cell Shape Classifier (https://www.mathworks.com/matlabcentral/fileexchange/37497-cell-shape-classifier), MATLAB Central File Exchange. 检索来源 .
MATLAB 版本兼容性
平台兼容性
Windows macOS Linux类别
- Sciences > Biological and Health Sciences > Biomedical Imaging >
- AI, Data Science, and Statistics > Statistics and Machine Learning Toolbox > Classification >
标签
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!