Automatic Extraction of Salient Objects using Feature Maps

Identification and ranking of relevant image content

“…Similar approach was proposed to identify quantification of DNA damage in cells as ROI [17]. importance is proposed in [21]. Salient objects are extracted by applying a segmentation algorithm on a combination of image edge and color maps.…”

Unsupervised salient object segmentation from color images

“…The second class is the so-called salient objects, which attract human visual attention and can be any class of object. Due to its unsupervised and universal property, salient object segmentation [10][11][12][13][14][15][16] is more applicable to a broader range of applications.…”

“…Interactive object segmentation approaches [1][2][3][4] reduce the user interaction work by allowing users to simply mark the interested objects with scribbles, bounding rectangle, or contour points, and can obtain the interested objects with user-desired quality by further user's refinement. Compared with the former two ways, unsupervised segmentation approaches [5][6][7][10][11][12][13][14][15][16] do not need any user interaction work, and this is an obvious advantage for implementation in some applications. Semantic objects segmented by unsupervised segmentation approaches can be categorized to two classes.…”

“…In [11], a support vector machine is trained to select salient regions for clustering into the salient object. In [12], salient object is extracted by performing analysis on a combination map that integrates color and edge feature maps.…”

Unsupervised foreground segmentation using background elimination and graph cut techniques

“…Introduction: Foreground segmentation is the task of identifying objects from the background [1,2]. To effectively extract the foreground surrounded by backgrounds, it is necessary to distinguish the background regions from still images.…”