This paper presents a novel method for the multi-scale segmentation and shape characterization of unsharp, blobby structures in gray-valued image data. The iso-level curves of unsharp blobs depict a set of concentric curves, representing the shape by the geometric properties of the level curves and by the radial intensity function. Applying Euclidean shortening flow progressively smoothes the level-curves and lets them converge to circles before they disappear at singularity points. We interleave object detection with shape computation by analyzing the continuous extremum paths created by singularities in scale-space. Assuming radially symmetric image structures, such singularity traces attributed with evolution time and gray level fully represent the shape information and allow reconstruction and quantitative evaluation. More complex geometry of simple, closed curves can be considered by analyzing the set of level curves associated with each trace. The project is driven by the medical problem of detecting white matter lesions in medical images. Here, a complete characterization of the gray-level structure represents a crucial descriptive feature for monitoring the progress of disease. The 2-D scheme is finally extended to the evolution of level surfaces in volume data, addressing the remaining open issues.