Abstract
Image segmentation is a challenging task in the field of image processing and computer vision. In order to obtain an accurate segmentation performance, user interaction is always used in practical image-segmentation applications. However, a good segmentation method should not rely on much prior information. In this paper, an efficient superpixel-guided interactive image-segmentation algorithm based on graph theory is proposed. In this algorithm, we first perform the initial segmentation by using the MeanShift algorithm, then a graph is built by taking the pre-segmented regions (superpixels) as nodes, and the maximum flow–minimum cut algorithm is applied to get the superpixel-level segmentation solution. In this process, each superpixel is represented by a color histogram, and the Bhattacharyya coefficient is chosen to calculate the similarity between any two adjacent superpixels. Considering the over-segmentation problem of the MeanShift algorithm, a narrow band is constructed along the contour of objects using a morphology operator. In order to further segment the pixels around edges accurately, a graph is created again for those pixels in the narrow band and, following the maximum flow–minimum cut algorithm, the final pixel-level segmentation is completed. Extensive experimental results show that the presented algorithm obtains much more accurate segmentation results with less user interaction and less running time than the widely used GraphCut algorithm, Lazy Snapping algorithm, GrabCut algorithm and a region merging algorithm based on maximum similarity (MSRM).
Highlights
We present an efficient superpixel-guided interactive image-segmentation algorithm based on graph theory
As can be clearly seen from the table, compared with the number of pixels, the number of superpixels processed by the pre-segmentation techniques has been greatly reduced, where we can see that the MeanShift algorithm is more effective than the WaterShed algorithm
The parameters of our method were set as follows: the size of the color histogram was set as Z = 4096, the balance control parameter in the graph model was set as λ = 1.0, the size of the square structural element in the morphology process was set as b = 2, and the number of Gaussians of
Summary
Image segmentation, which aims to extract objects of interest from a complex background for object detection, tracking, recognition, scene analysis, etc., is one of the basic problems in image processing, and has been widely used in pattern recognition and computer vision [1,2,3,4,5,6,7]. MeanShift algorithm is applied to pre-segment an image into regions (superpixels), and the proposed algorithm constructs a weighted directed graph whose nodes are composed of the pre-partitioned regions This model considers the correlation between adjacent Symmetry superpixels, alsoREVIEW takes the relationship between each superpixel and the. Bhattacharyya coefficients to algorithm measureconstructs the similarity between superpixels, and intowe regions (superpixels), and the proposed a weighted directed graph the maximum cutofalgorithm is performed to model obtainnot theonly first-stage whose flow–minimum nodes are composed the pre-partitioned regions This considerssegmentation the correlation between adjacent and takes the relationship between each superpixelthe andproposed results.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
