A method for segmenting an image includes receiving one or more object seeds associated with one or more first object portions represented by a first image and receiving one or more background seeds associated with one or more first background portions represented by the first image. A second image representing one or more second object portions and one or more second background portions is received, wherein the first image comprises a plurality of subportions and the second image is a subportion of the plurality of subportions of the first image. The second image is processed based on the one or more object seeds and the one or more background seeds to determine the one or more second object portions and the one or more second background portions.