Improved Image Segmentation via Cost Minimization of Multiple Hypotheses
This work addresses the problem of parameter selection in image segmentation for computer vision applications, offering an incremental improvement over existing methods.
The paper tackles the challenge of selecting parameters for image segmentation to achieve human-like perceptual grouping by generating multiple segmentation hypotheses from different parameter choices and using a cost minimization framework to select stable segments that describe natural contours. It shows improved results compared to state-of-the-art algorithms and robustness to the choice of segmentation kernel.
Image segmentation is an important component of many image understanding systems. It aims to group pixels in a spatially and perceptually coherent manner. Typically, these algorithms have a collection of parameters that control the degree of over-segmentation produced. It still remains a challenge to properly select such parameters for human-like perceptual grouping. In this work, we exploit the diversity of segments produced by different choices of parameters. We scan the segmentation parameter space and generate a collection of image segmentation hypotheses (from highly over-segmented to under-segmented). These are fed into a cost minimization framework that produces the final segmentation by selecting segments that: (1) better describe the natural contours of the image, and (2) are more stable and persistent among all the segmentation hypotheses. We compare our algorithm's performance with state-of-the-art algorithms, showing that we can achieve improved results. We also show that our framework is robust to the choice of segmentation kernel that produces the initial set of hypotheses.