Seed segmentation using l
View Patent ↗A computer readable medium embodying instructions executable by a processor to perform a method for seeded image segmentation using l ∞ minimization, the method including providing an image comprising a set of pixels, wherein a foreground seed label is given for at least one pixel of the image and a background seed label is given for at least another pixel of the image, determining an affinity function for every pair of neighbor pixels, solving the l ∞ minimization, which assigns a probability to each pixel, labeling each pixel as foreground pixel or background pixel according to a threshold of the probability, and outputting the image including the segmentation labels.
1. A non-transitory computer readable medium embodying instructions executable by a processor to perform a method for seeded image segmentation using l ∞ minimization, the method steps comprising:
providing an image comprising a set of pixels, wherein a foreground seed label is given for at least one pixel of the image and a background seed label is given for at least another pixel of the image;
determining an affinity function for every pair or neighbor pixels by the l ∞ minimization;
solving the l ∞ minimization, which assigns a probability to each pixel;
labeling each pixel as foreground pixel or background pixel according to a threshold of the probability; and
outputting the image including the segmentation labels.
2. The method of claim 1 , wherein the threshold is 0.5.
3. The method of claim 1 , further comprising optimizing the probability according to a shortest path between the foreground seed label and the background seed label.
4. The method of claim 1 , further comprising an l 2 regularization of the l ∞ minimization.
5. The method of claim 1 , wherein the l ∞ minimization is written as
min max i ij εE w ij |x i −x j |s.t. x F =1, x B =0
wherein w ij is a weighting function, e ij is an edge between neighboring pixels i and j of a set E, and x is the probability of a foreground pixel F and a background pixel B.
6. The method of claim 1 , wherein the outputting comprises at least one of displaying the image including the segmentation labels and storing the image including the segmentation labels to a computer readable medium.
7. A system for image segmentation comprising:
a source of an image to be segmented;
an input device for receiving foreground and background seed pixels of the image to be segmented;
a processor executing a instructions embodied in a computer readable medium for seeded image segmentation of the image using l ∞ minimization and determining a segmentation of the image identifying a foreground portion and a background portion of the image; and
a output device for receiving the segmentation determined by the processor, wherein the l ∞ minimization is written as
min max i ij εE w ij |x i −x j |s.t. x F =1, x B =0
wherein w ij is a weighting function, e ij is an edge between neighboring pixels i and j of a set E, and x is the probability of a foreground pixel F and a background pixel B.
8. The system of claim 7 , wherein the output device is one of a display for displaying the segmentation determined by the processor and the computer readable medium, wherein the computer readable medium stores the segmentation determined by the processor.