Iterative method for salient foreground detection and multi-object segmentation
A system and method that performs iterative foreground detection and multi-object segmentation in an image is disclosed herein. A new background prior is introduced to improve the foreground segmentation results. Three complimentary methods detect and segment foregrounds containing multiple objects. The first method performs an iterative segmentation of the image to pull out the salient objects in the image. In a second method, a higher dimensional embedding of the image graph is used to estimate the saliency score and extract multiple salient objects. A third method uses a metric to automatically pick the number of eigenvectors to consider in an alternative method to iteratively compute the image saliency map. Experimental results show that these methods succeed in accurately extracting multiple foreground objects from an image.
1. A method for detecting and segmenting multiple foreground objects, comprising:
(a) performing, on a digital image, an initial foreground segmentation using superpixel segmentation with an augmented background model to generate an image adjacency graph;
(b) computing a set of saliency scores from the image adjacency graph;
(c) finding a set of nodes or superpixels from the image adjacency graph for which the saliency score is greater than a specified threshold;
(d) modifying the image adjacency graph by cutting out nodes in the set of nodes and storing the saliency scores of these nodes;
(e) determining new saliency scores for a region that remains in the image adjacency graph by computing the Fiedler Vector of the modified image adjacency graph;
(f) combining the new saliency scores of the region with the saliency scores for the nodes from the set of nodes to obtain a new saliency image;
(g) computing a new overall saliency score for the new saliency image;
repeating steps (c) to (g) a predetermined number of iterations; and
choosing a segmentation map with highest overall image saliency score.
2. The method of claim 1 wherein the augmented background model comprises a multi-color background model obtained by clustering the superpixel colors represented in a Lab color space.
3. The method of claim 1 , wherein the image adjacency graph comprises a reduced image representation of the digital image, in which a group of pixels are represented by an average color of pixels in a corresponding superpixel and represented by a node in a graph, and local relationships in the image are represented by connecting two regions in the graph if the corresponding superpixels share a border in the original image.
4. The method of claim 1 wherein all saliency scores are computed by using the Fiedler vector, and rescaling the resulting scores between [0; 1].
5. The method of claim 1 , wherein the new overall saliency score is computed by combining a silhouette score and a mean image saliency.
6. The method of claim 1 wherein cutting out nodes comprises removing the nodes from a graph representation of the digital image that have saliency values exceeding a predetermined threshold.