Watershed Segmentation
Image segmentation - Watershed More...
Detailed Description
Image segmentation - Watershed
In the study of image processing, a watershed is a transformation defined on a grayscale image. The name refers metaphorically to a geological watershed, or drainage divide, which separates adjacent drainage basins. The watershed transformation treats the image it operates upon like a topographic map, with the brightness of each point representing its height, and finds the lines that run along the tops of ridges.
- See also
- Watershed - Image segmentation
Collaboration diagram for Watershed Segmentation:
Functions | |
| template<class T , class labelT > | |
| RES_T | basins (const Image< T > &imIn, const Image< labelT > &imMarkers, Image< labelT > &imBasinsOut, const StrElt &se=DEFAULT_SE) |
| Constrained basins. More... | |
| template<class T , class labelT > | |
| RES_T | watershed (const Image< T > &imIn, const Image< labelT > &imMarkers, Image< T > &imOut, Image< labelT > &imBasinsOut, const StrElt &se=DEFAULT_SE) |
| Constrained watershed. More... | |
| template<class T > | |
| RES_T | lblSkiz (Image< T > &labelIm1, Image< T > &labelIm2, const Image< T > &maskIm, const StrElt &se=DEFAULT_SE) |
| Skiz on label image. More... | |
| template<class T1 , class T2 > | |
| RES_T | inflBasins (const Image< T1 > &imIn, Image< T2 > &basinsOut, const StrElt &se=DEFAULT_SE) |
| Influences basins. More... | |
| template<class T > | |
| RES_T | inflZones (const Image< T > &imIn, Image< T > &imOut, const StrElt &se=DEFAULT_SE) |
| Influences zones. More... | |
| template<class T > | |
| RES_T | waterfall (const Image< T > &gradIn, const Image< T > &wsIn, Image< T > &imGradOut, Image< T > &imWsOut, const StrElt &se=DEFAULT_SE) |
| Waterfall. | |
| template<class T > | |
| RES_T | waterfall (const Image< T > &gradIn, UINT nLevel, Image< T > &imWsOut, const StrElt &se=DEFAULT_SE) |
| Waterfall. | |
Function Documentation
◆ basins()
| RES_T smil::basins | ( | const Image< T > & | imIn, |
| const Image< labelT > & | imMarkers, | ||
| Image< labelT > & | imBasinsOut, | ||
| const StrElt & | se = DEFAULT_SE |
||
| ) |
Constrained basins.
Hierachical queue based algorithm as described by S. Beucher (2011) [1]
- Parameters
-
[in] imIn Input image. [in] imMarkers Label image containing the markers. [out] imBasinsOut (optional) Output image containing the basins. [in] se Structuring element After processing, this image will contain the basins with the same label values as the initial markers.
Example
from smilPython import *
# Load an image
imIn = Image("https://smil.cmm.minesparis.psl.eu/images/DNA_small.png")
imIn.show()
# Create a gradient image
imGrad = Image(imIn)
gradient(imIn, imGrad)
# Manually impose markers on image
imMark = Image(imIn, "UINT16")
imMark << 0
# One for the background...
imMark.setPixel(75, 40, 1)
# and one on two connected particules
imMark.setPixel(78, 86, 2)
imMark.setPixel(88, 76, 3)
# Dilate the markers to avoid to be blocked in a minimum
dilate(imMark, imMark, 2)
#imMark.showLabel()
# Create the watershed
imWS = Image(imIn)
watershed(imGrad, imMark, imWS)
# Display output
imWS.show()
# Display the output as overlay on the original image
imIn.getViewer().drawOverlay(imWS & 1)
RES_T drawOverlay(const Image< T > &imToDraw, Image< T > &imOut)
Draw overlay.
Definition: DImage.hpp:516
RES_T dilate(const Image< T > &imIn, Image< T > &imOut, const StrElt &se=DEFAULT_SE, const T borderVal=ImDtTypes< T >::min())
dilate() - Morphological grayscale dilation
Definition: DMorphoBase.hpp:65
RES_T gradient(const Image< T > &imIn, Image< T > &imOut, const StrElt &se=DEFAULT_SE)
gradient() - Morphological gradient
Definition: DMorphoResidues.hpp:62
RES_T watershed(const Image< T > &imIn, const Image< labelT > &imMarkers, Image< T > &imOut, Image< labelT > &imBasinsOut, const StrElt &se=DEFAULT_SE)
Constrained watershed.
Definition: DMorphoWatershed.hpp:402
◆ watershed()
| RES_T smil::watershed | ( | const Image< T > & | imIn, |
| const Image< labelT > & | imMarkers, | ||
| Image< T > & | imOut, | ||
| Image< labelT > & | imBasinsOut, | ||
| const StrElt & | se = DEFAULT_SE |
||
| ) |
Constrained watershed.
Hierachical queue based algorithm as described by S. Beucher (2011) [1]
- Parameters
-
[in] imIn Input image. [in] imMarkers Label image containing the markers. [in] se Structuring element [out] imOut Output image containing the watershed lines. [out] imBasinsOut (optional) Output image containing the basins. After processing, this image will contain the basins with the same label values as the initial markers.
Example
from smilPython import *
# Load an image
imIn = Image("https://smil.cmm.minesparis.psl.eu/images/DNA_small.png")
imIn.show()
# Create a gradient image
imGrad = Image(imIn)
gradient(imIn, imGrad)
# Manually impose markers on image
imMark = Image(imIn, "UINT16")
imMark << 0
# One for the background...
imMark.setPixel(75, 40, 1)
# and one on two connected particules
imMark.setPixel(78, 86, 2)
imMark.setPixel(88, 76, 3)
# Dilate the markers to avoid to be blocked in a minimum
dilate(imMark, imMark, 2)
#imMark.showLabel()
# Create the watershed
imWS = Image(imIn)
watershed(imGrad, imMark, imWS)
# Display output
imWS.show()
# Display the output as overlay on the original image
imIn.getViewer().drawOverlay(imWS & 1)
◆ lblSkiz()
| RES_T smil::lblSkiz | ( | Image< T > & | labelIm1, |
| Image< T > & | labelIm2, | ||
| const Image< T > & | maskIm, | ||
| const StrElt & | se = DEFAULT_SE |
||
| ) |
Skiz on label image.
Performs the influence zones on a label image as described by S. Beucher (2011) [3] If a maskIm is provided, the skiz is geodesic.
◆ inflBasins()
| RES_T smil::inflBasins | ( | const Image< T1 > & | imIn, |
| Image< T2 > & | basinsOut, | ||
| const StrElt & | se = DEFAULT_SE |
||
| ) |
Influences basins.
Performs the influence basins using the lblSkiz function. Input image is supposed to be binary.
