Thinning grayscale well-composed images
Introduction
A digital image may be seen as the digitization of a piecewise continuous function. The discontinuities of this function are of primary importance in many shape recognition processes, as they usually describe the shapes of objects appearing on an image.
The notion of discontinuity is lost once a function is digitized. The structure describing domains where the underlying piecewise function is continuous is called a segmentation. Two main dual approaches to segmentation may be distinguished. The first approach consists of approximating the discrete function by a piecewise continuous function, and is usually referred to a region oriented segmentation. The second approach tries to directly catch the discontinuities of an underlying continuous function and is referred to a contour oriented segmentation. A topological segmentation may be viewed in this context as a process capturing both domains where an underlying piecewise function is continuous, and the set of points describing the discontinuities of the function.
A piecewise continuous function may be represented by a topological map, which can be viewed as a partition of the plane into three sets of points: a finite set S of points, a finite set A of disconnected Jordan arcs having elements of S as extremities, and a set of connected domains, the faces, whose boundaries are unions of elements of S and A. In this paper we aim to construct a topological segmentation that has a topological map structure consistent with a digital topological framework and with a topological map representing a continuous piecewise function.
Recent works (Braquelaire and Brun, 1998; Fiorio, 1996) aim to develop a coherent topological structure describing a digital image from the information provided by regions. The different structures described use a discrete topology based on the decomposition of the support domain of an image into three kinds of elements of different dimensions, i.e. surface elements, associated with the discrete points of the support, edge elements, which are the edges separating two surface elements, and vertices of the so defined grid (Fiorio, 1995; Kovalevsky, 1989). On one hand, such a partition has nice topological properties, but on the other hand, it suffers from many practical drawbacks, such as the amount of memory needed to store the entire partition, and difficulties faced when trying to construct the partition from the contour information.
Alternatively, a topological partition of an image may be directly defined by a digital topology involving only points on the square grid. In order to face the connectivity paradox, several neighborhood systems are usually used together. This is done either by considering different adjacency relations for points belonging to a set and its complement (Kong and Rosenfeld, 1989), or by assigning different neighborhoods to each point of in a data independent manner, which can be formally stated using the framework proposed in (Khalimsky et al., 1990) (this framework can also be used in the cellular complex approach (Kovalevsky, 1989)).
Watersheds or more generally graytone skeletons can be used in this context to retrieve a vertex/arc network (crest lines) and faces (catchment basins) from a discrete topographic surface such as the modulus of the gradient viewed as a relief, which is exactly the sought for partition of the image.
However, many consistency problems are encountered on a square grid. Approaches that work by suppressing points from a potential crest network (grayscale thinning) or by adding points to connected sets of points (Arcelli, 1981; Bertrand et al., 1997) do not usually guarantee that the extracted crest network is thin (Fig. 1). Thick configurations of crests pose obvious problems when one is trying to link points from the resulting crests network in order to obtain digital curves and vertices of the topological partition. Approaches that work by linking potential crest points (Meyer, 1989; Pierrot Deseilligny et al., 1998), constructing a raster graph, do not usually guarantee that the faces defined by the cycles of the graph are composed of a unique connected component (Fig. 2).
Latecki proposed to face the problem of thickness of skeletons on digital binary images by forbidding some configurations of points (Latecki et al., 1995). He proposed a thinning operator that preserves the properties of the so called well-composed binary images, resulting in a thin skeleton. He also demonstrated a Jordan theorem that is verified on well-composed sets of points. He extended the property of well-composedness to multicolor images.
In this contribution, we first recall some classical notions of digital topology (Section 2) and some properties of well-composed sets (Section 3). We redefine grayscale well-composed images using the cross section topology formalism (Bertrand et al., 1996, Bertrand et al., 1997; Meyer, 1989), adapt a grayscale thinning algorithm to well-composed graylevel images, and prove some of its properties (Section 4). We derive an algorithm which constructs a topological partition from an irreductible thin well-composed image, and finally present an application (Section 5).
Section snippets
Digital topology: basic notions
A discrete image I is a function from to a set E. If E={0,1}, I is said to be a binary digital image. If E={0,…,k}, I is said to be a grayscale digital image. A point of a digital image is a couple . Two points and are
- •
4-adjacent if and only if d4(p1,p2)=|x1−x2|+|y1−y2|=1.
- •
8-adjacent if and only if d8(p1,p2)=max(|x1−x2|,|y1−y2|)=1.
The n-neighborhood Γn(p) of a point p is the set of all the points n-adjacent to p, with n=4 or n=8. A point p is said to be
Well-composed sets of points
A coherent topological structure of a digital image should respect an equivalent of the Jordan curve theorem by which the complement of a simple closed curve is a set composed of two connected components. One of the major drawbacks of digital topology is known as the connectivity paradox. If the same neighborhood system is used for studying the connectivity of a set and its complement, the Jordan curve theorem does not have its digital counterpart on a square grid. In order to solve this
Well-composed grayscale images
Meyer (1989) proposed to characterize crest points on a grayscale image by studying a family of binary images, which are obtained by thresholding the original image by all possible threshold values. Crest points are the points that change the homotopy of one of the binary images. More recently, cross-section topology has been introduced by Bertrand et al., 1996, Bertrand et al., 1997. It has led to the development of an efficient watershed algorithm (Couprie and Bertrand, 1997). A definition of
Constructing a topological partition and application
The topological map defined on the digital plane is composed of:
- •
a set S of points which are vertices of the map,
- •
a set A of arcs, which are 4-connected disconnected digital curves, of which the extremities are elements of S,
- •
a set F of faces, which are 4-connected sets of points, of which the boundaries are elements of S and A.
Such a structure can be built from an irreducible well-composed grayscale image without peak, obtained by a leveling transformation of a well-composed grayscale image. The
Conclusion
In this contribution, we have used the cross section topology formalism in order to define a thinning operator which conserves well-composedness on gray level images. Moreover, we have proposed a way to construct a topological map from the resulting thin image, and have shown that the obtained map is coherent in the sense that:
- •
a bijection exists between cycles of the map and 4-connected regions of the thin image,
- •
each Jordan arc of the map is a 4-connected digital curve.
The map can be
References (16)
Pattern thinning by contour tracing
Comput. Graph. Image Process.
(1981)- et al.
Image segmentation with topological maps and inter-pixel representation
J. Vis. Commun. Image Represent.
(1998) - et al.
Computer graphics and connected topologies on finite ordered sets
Topology and its Applications
(1990) Finite topology as applied to image analysis
Computer Vision, Graphics and Image Processing
(1989)- et al.
Digital topology, introduction and survey
Comput. Vis. Graph. Image Process.
(1989) Multicolor well-composed pictures
Pattern Recogn. Lett.
(1995)- et al.
Well-composed sets
Comput. Vis. Image Understand.
(1995) Skeletons and perceptual graphs
Signal Process.
(1989)
Cited by (17)
3D well-composed polyhedral complexes
2015, Discrete Applied MathematicsCitation Excerpt :3D well-composed images [12] enjoy important topological and geometrical properties in such a way that several algorithms used in computer vision, computer graphics and image processing are simpler. For example, thinning algorithms can be simplified and naturally made parallel if the input image is well-composed [14,16]; some algorithms for computing surface curvature or extracting adaptive triangulated surfaces assume that the input image is well-composed [10]. However, our main motivation is that of (co)homology computations on the cell complex representing the 3D image [3,5].
Strong Euler well-composedness
2022, Journal of Combinatorial OptimizationTopological Properties of the First Non-Local Digitally Well-Composed Interpolation on n-D Cubical Grids
2020, Journal of Mathematical Imaging and VisionA Tutorial on Well-Composedness
2018, Journal of Mathematical Imaging and VisionHow to make nD images well-composed without interpolation
2015, Proceedings - International Conference on Image Processing, ICIPTopology-preserving rigid transformation of 2D digital images
2014, IEEE Transactions on Image Processing