Interactive Image Segmentation Based on Region Merging Using Hierarchical Match Mechanism

Image segmentation is to segment an image into disjoint parts. It is a fundamental and challenging task in computer vision. We present an interactive image segmentation method based on region merging. An image is first over-segmented into many super-pixels using bottom-up methods. Then some strokes are manually labeled on the over-segmented image. The strokes give information about object classes occurring in the image. Low-level features of all the super-pixels are extracted. We propose a hierarchical match mechanism to seek similar pairs. The hierarchical match mechanism is exploited in the merging process to improve segmentation performance. Experiments are performed and acceptable results are provided.