The MontiVision Segmentation filter is used for gray or color image segmentation. Manual and automatic thresholding is supported. Either a single color channel is specified for the segmentation or all channels are used.
| File: | binary.ax |
| CLSID: | CLSID_MVBinary |
Supported Interfaces
Supported COM Events
The filter supports following events through the IMVFilterEvents interface. Look at the filter events page for more information.
| Event |
Value |
Description |
| MV_ALGORITHM_START_EVENT |
0 |
A filter starts processing a video frame or sample. |
| MV_ALGORITHM_END_EVENT |
1 |
A filter finishes processing a video frame or sample. |
Supported I/O Pins
| I/O Pin | Corresponding method | Discription |
|
Threshold In [Byte] |
IMVBinary::put_Threshold | Sets the threshold used by the fixed threshold algorithm. |
|
Threshold Out [Byte] |
IMVBinary::get_Threshold | Returns the currently used fixed or automatic threshold, depending on the currently selected binarization method. |
Attention:
I/O Pins are dominat, what means that they have a higher
priority than the corresponding method of the filters
interface!
MV Segmentation Property Page
Implements the IMVBinary configuration interface.
| Manual Threshold | The segmentation threshold is set manually using the slider. |
| Entropy (Johannsen) | Automatic thresholding using the Johannsen entropy method. |
| Edge Pixels (Weska) | Automatic thresholding using the Weszka edge pixel method. The laplacian of the frame is calculated. Pixels that have a laplacian greater than 85% of the highest laplacian are used to create a histogram. The threshold is selected using the Two Peaks algorithm (see below). |
| Two Peaks | Automatic thresholding using the Johannsen entropy method. Most histograms contain two peaks, one represents the foreground and one the background. The threshold is selected using the smallest value between the peaks. |
| Color Channel | Used to select the channel that is used for segmentation. |