Hyperconnected Attribute Filters Based on k-Flat Zones

Ouzounis, Georgios K.; Wilkinson, Michael H. F.

doi:10.1109/tpami.2010.74

Cited by 37 publications

(42 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[1] for more details on connectivity and [57][58][59][60] for examples of advanced connectivities) determines the set of edges E connecting those pixels and associates an undirected graph G = (V, E) to the image. An edge between two adjacent pixels p and q is denoted by e p,q or e q,p .…”

Section: Images As Graphsmentioning

confidence: 99%

Partition and Inclusion Hierarchies of Images: A Comprehensive Survey

2018

View full text Add to dashboard Cite

Abstract:The theory of hierarchical image representations has been well-established in Mathematical Morphology, and provides a suitable framework to handle images through objects or regions taking into account their scale. Such approaches have increased in popularity and been favourably compared to treating individual image elements in various domains and applications. This survey paper presents the development of hierarchical image representations over the last 20 years using the framework of component trees. We introduce two classes of component trees, partitioning and inclusion trees, and describe their general characteristics and differences. Examples of hierarchies for each of the classes are compared, with the resulting study aiming to serve as a guideline when choosing a hierarchical image representation for any application and image domain.

show abstract

Section: Images As Graphsmentioning

confidence: 99%

Partition and Inclusion Hierarchies of Images: A Comprehensive Survey

2018

View full text Add to dashboard Cite

show abstract

“…In particular, we can distinguish crisp and fuzzy clustering. (20) is a crisp attribute cluster lter if:…”

Section: The Clustering Approachmentioning

confidence: 99%

“…This is achieved ) and volume(λ = ) (e) attribute cluster lter using 5 attributes and k-means(k = ) (e) k-flat ltering(k = ) using both fuzzy c-means and k-means with k = using 5 attributes of surface area, X-extent, Y-extent, Z-extent, volume. The same result is only matched using hyperconnectivity [20] Fig. 5(f), with hyperconnectivity k-at zones [20], which are connected regions of maximal extent, where the total grey level variation is not more than k. This restriction to grey-level range automatically restricts the size to which the regions can grow, yielding overlapping pseudo-at zones, which improves the enhancement of internal details.…”

Section: Ct-knee Volumementioning

confidence: 99%

“…E cient techniques and algorithms have been developed for extracting image components that are useful in the representation and description of shapes. These techniques have found application in medical imaging [8,18,38], document analysis [20], video processing [24], content-based compression [33], color processing [32] and remote sensing applications [42].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Cluster Based Vector Attribute Filtering

Kiwanuka

Wilkinson

2016

Mathematical Morphology - Theory and Applications

View full text Add to dashboard Cite

Morphological attribute lters operate on images based on properties or attributes of connected components. Until recently, attribute ltering was based on a single global threshold on a scalar property to remove or retain objects. A single threshold struggles in case no single property or attribute value has a suitable, usually multi-modal, distribution. Vector-attribute ltering allows better description of characteristic features for 2D images. In this paper, we apply vector-attribute ltering to 3D and incorporate unsupervised pattern recognition, where connected components are classi ed based on the similarity of feature vectors. Using a single attribute allows multi-thresholding for attribute lters where more than two classes of structures of interest can be selected. In vector-attribute lters automatic clustering avoids the need for either setting very many attribute thresholds, or nding suitable class prototypes in 3D and setting a dissimilarity threshold. Explorative visualization reduces to visualizing and selecting relevant clusters. We show that the performance of these new lters is better than those of regular attribute lters in enhancement of objects in medical images.

show abstract

“…Additionally, data transformation or mapping functions [6] may be added as a constituent in such a framework. Utilizing mapping functions falls within the context of defining connected components more elaborately, also considered for connected component segmentation algorithms via additional processing (prefiltering/post-filtering) [12,13], via analyzing scene-wide statistics [14,15], and via the notion of hyperconnections [16][17][18]. Adding mapping functions into such a framework results in three distinct, interdependent parameter constituents that need consideration.…”

Section: Introductionmentioning

confidence: 99%

On Attribute Thresholding and Data Mapping Functions in a Supervised Connected Component Segmentation Framework

Fourie¹

2015

Remote Sensing

View full text Add to dashboard Cite

Search-centric, sample supervised image segmentation has been demonstrated as a viable general approach applicable within the context of remote sensing image analysis. Such an approach casts the controlling parameters of image processing-generating segments-as a multidimensional search problem resolvable via efficient search methods. In this work, this general approach is analyzed in the context of connected component segmentation. A specific formulation of connected component labeling, based on quasi-flat zones, allows for the addition of arbitrary segment attributes to contribute to the nature of the output. This is in addition to core tunable parameters controlling the basic nature of connected components. Additional tunable constituents may also be introduced into such a framework, allowing flexibility in the definition of connected component connectivity, either directly via defining connectivity differently or via additional processes such as data mapping functions. The relative merits of these two additional constituents, namely the addition of tunable attributes and data mapping functions, are contrasted in a general remote sensing image analysis setting. Interestingly, tunable attributes in such a context, conjectured to be safely useful in general settings, were found detrimental under cross-validated conditions. This is in addition to this constituent's requiring substantially greater computing time. Casting connectivity definitions as a searchable component, here via the utilization of data mapping functions, proved more beneficial and robust in this context. The results suggest that further investigations into such a general framework could benefit more from focusing on the aspects of data mapping and modifiable connectivity as opposed to the utility of thresholding various geometric and spectral attributes.

show abstract

Hyperconnected Attribute Filters Based on k-Flat Zones

Cited by 37 publications

References 46 publications

Partition and Inclusion Hierarchies of Images: A Comprehensive Survey

Partition and Inclusion Hierarchies of Images: A Comprehensive Survey

Cluster Based Vector Attribute Filtering

On Attribute Thresholding and Data Mapping Functions in a Supervised Connected Component Segmentation Framework

Contact Info

Product

Resources

About