Abstract
The low intensity of several pixels can be a crucial problem for the segmentation of moving object because the object cannot be separated perfectly from the background. This paper proposes Gabor filtering to enhance the intensity of foreground in automatic segmentation of moving object. The experiments evaluate the application of Gabor filter and Fuzzy C-Means (FCM) clustering for segmentation of moving objects showing that FCM with Gabor filter is able to perform a good differentiation process of foreground and background which is better than FCM with no Gabor filter. In the work, several distance metrics have been evaluated. The result shows that Minkowski distance is more accurate than the other distances.
Original language | English |
---|---|
Pages (from-to) | 1054-1060 |
Number of pages | 7 |
Journal | International Review on Computers and Software |
Volume | 10 |
Issue number | 10 |
DOIs | |
Publication status | Published - 2015 |
Keywords
- Distance measure
- Fuzzy C-means
- Gabor filter
- Segmentation for moving object