Hierarchical segmentation

Chini M, Chiancone A, Stramondo S (2014) Scale object selection (SOS) through a hierarchical segmentation by a multi-spectral per-pixel classification. Pattern Recognit Lett 49 214-223... 

A similar way is proposed in the work of Hutchinson and Chen 2005. Here the object selection is performed through a hierarchical segmentation procedure. 

The aim of this chapter is to introduce an approach towards geological modeling which is able to encompass the structure dictated by sequence stratigraphy, and fulfill modeling constraints from a wide range of disciplines. Motivated by the hierarchical structuring of depositional sequences, a hierarchical segmentation approach is presented. 

E. Monsen, T. Randen, L. Sqnneland, and J.E. Odegard (2004) Geological model building a hierarchical segmentation approach. This volume. 

A hierarchical system is the simplest type ofdatabase system. In this form, the var-iou.s data typc.s also called entities (sec figure 5-,3) arc as.signcd. systematically to various levels (Figure 5-5). The hierarchical system is represented as an upside-down tree with one root segment and ordered nodes. Each parent object can have one or more children (objects) but each child has only one parent. If an object should have more than one parent, this entity has to be placed a second time at another place in the database system. 

In order to trace (find, change, add, or delete) a segment in the database, the sequence in which the data arc read is important. Thus, the sequence of the hierarchical path is parent > child > siblings. The assignment of the data entities uses pointers. In our example, the hierarchical path to K is traced in Figure 5-fi. 

A systematic analysis of a process signal over (1) different segments of its time record and (2) various ranges of frequency (or scale) can provide a local (in time) and multiscale hierarchical description of the signal. Such description is needed if an intelligent computer-aided tool is to be con--structed in order to (1) localize in time the step and spike from the equipment faults (Fig. 1), or the onset of change in sensor noise characteristics, and (2) extract the slow drift and the periodic load disturbance. 

Despite the complexity of the decay functions generated from consideration of the hierarchical motion of the chain segments, a more phenomenological approach has met with some success. In concentrated solution (C > C ), or in the molten state when the molar mass is greater than the critical mass for entanglements, there is often observed two separate decays with largely-different time constants of decay ... 

On the other hand, the apparatus is so cheap and simple to use that it may be feasible to run several independent systems in parallel. The obvious analogy is between a hierarchical computer system linked to a mainframe and a series of dedicated microcomputers. Also the cheapness, speed and simplicity of non-segmented systems make them attractive for a small number of determinations, less than would be considered viable with Autoanalysers. 

The readout by STEM will be achieved in either of the two ways. First, hierarchical automated particle selection and segmentation algorithms will... 

The phase behaviour of biomimetic polypeptide-based copolymers in solution was described and discussed with respect to the occurrence of secondary structure effects. Evidently, incorporation of crystallisable polypeptide segments inside the core of an aggregate has impact on the curvature of the corecorona interface and promotes the formation of fibrils or vesicles or other flat superstructures. Spherical micelles are usually not observed. Copolymers with soluble polypeptide segments, on the other hand, seem to behave like conventional block copolymers. A pH-induced change of the conformation of coronal polypeptide chains only affects the size of aggregates but not their shape. The lyotropic phases of polypeptide copolymers indicate the existence of hierarchical superstructures with ordering in the length-scale of microns. 

---