China 
Africa 
Explore chapters and articles related to this topic
Image Segmentation
Published in N.C. Basantia, Leo M.L. Nollet, Mohammed Kamruzzaman, Hyperspectral Imaging Analysis and Applications for Food Quality, 2018
Sylvio Barbon, Ana Paula Ayub da Costa Barbon, N.A. Valous, D.F. Barbin
Image segmentation when based on mathematical morphology deals with fundamental operations on two sets: one set is used to satisfy the other set having a finely determined shape and size, known as structuring element (SE) (Serra, 1982). The structural matching of elements uses the mathematical morphological operations of erosion and dilation, which can be graphically illustrated by viewing the image data as an imaginary topographic relief where higher elevations are represented by brighter tones. Thus, the SE defines the new (dilated or eroded) scene based on its spatial properties such as height or width as a result of sliding the SE over the topographical relief. The application of morphological operators to hyperspectral images, which are based on multilayered data with hundreds of spectral channels, is not straightforward. A straightforward approach starts with grayscale morphology techniques for each channel separately (marginal approach). It is often unacceptable, because it may result in new spectral constituents that are not present in the original hyperspectral image as a result of processing the channels separately (Comer and Delp, 1999). An alternative (and perhaps more appropriate) way to approach the problem of multichannel morphology is to treat the data at each pixel as a vector (Soille, 2003). It is important to define an appropriate combination of vectors in vector space, which could be either spectral-domain partitioning or spatial-domain partitioning (Plaza et al., 2009).
Feature Extraction with Statistics and Decision Science Algorithms
Published in Ni-Bin Chang, Kaixu Bai, Multisensor Data Fusion and Machine Learning for Environmental Remote Sensing, 2018
In image processing, mathematical morphology is commonly used to examine interactions between an image and a set of structuring elements using certain operations, while the structuring element acts as a probe for extracting or suppressing specific structures of the image objects (Plaza, 2007). More specifically, morphological operations apply a structuring element to filter an image, while the value of each pixel in the output image is based on a comparison of the corresponding pixel in the input image with its neighbors. By choosing a proper size and shape of the neighborhood, a morphological operation that is sensitive to specific shapes in the input image can be constructed. The output of filtering process depends fully on the matches between the input image and the structuring element and the operation being performed (Quackenbush, 2004).
Studies on a formidable dot and globule related feature extraction technique for detection of melanoma from dermoscopic images
Published in Debatosh Guha, Badal Chakraborty, Himadri Sekhar Dutta, Computer, Communication and Electrical Technology, 2017
S. Chatterjee, D. Dey, S. Munshi
Morphological image processing has become a gold standard imaging toolbox and now a days, has been utilized in a wide range of industrial applications, biomedical image processing, pattern recognition, robot vision, etc. The foundation of mathematical morphology is the set theory (Soille 2004). Mathematical morphology deals with different geometrical structures of various shape and size. The aim of the different morphological operators is to extract relevant structures present in that image by probing the image with another set of structures small in size and known shape, referred to as SE. Choice of the SE is based on the prior knowledge of the relevant shapes and sizes of the structures present in that image. Erosion and dilation are the fundamental operations of mathematical morphology.
Water depth estimation from Sentinel-2 imagery using advanced machine learning methods and explainable artificial intelligence
Published in Geomatics, Natural Hazards and Risk, 2023
Vahideh Saeidi, Seyd Teymoor Seydi, Bahareh Kalantar, Naonori Ueda, Bahman Tajfirooz, Farzin Shabani
The two fundamental mathematical morphology operations are dilation and erosion (Licciardi et al. 2012). Erosion and dilation are the basic elements in the MP that are applied to a grayscale image with a specific structure element. This study involves a redesigning of Extended Morphological Profiles (EMP) based on a hierarchical morphological operator. In grayscale image analysis, each pixel’s intensity value is considered as the third dimension, corresponding to various features (Quackenbush 2004). Our proposed framework employs the output of a morphological layer in the individual stages as input for the next subsequent stages. Furthermore, the structure element size increases in each subsequent stage, using a 3 × 3, 5 × 5, and 7 × 7 kernel size filter. Figure 4 shows the general framework of the proposed feature extraction, consisting of three morphological erosion and dilation layers. The output of each layer is incorporated in the final output of the model, and the final output is constructed using the output of each layer in each stage
Novel Approach by Fuzzy Logic to Deal with Dynamic Analysis of Shadow Elimination and Occlusion Detection in Video Sequences of High-Density Scenes
Published in IETE Journal of Research, 2023
Hocine Chebi, Abdelkader Benaissa
The approaches old enough to analyze crowd behavior in video sequences generally include four essential steps: motion detection, segmentation, classification and tracking [6–9]. A shadow is a shady region created by the interposition of a dense object and a radiance source on the reflecting surface. Generally, shadows have a negative effect on the images or video sequences since its related main problem directs to a reduction or loss of image information. Indeed, the reduction of the image information conducts intensively to a distortion of the required parameters that are taken from the pixel values. The importance of shadow elimination resides then on the determination of the correct pixel values and on the estimation of the right detected region followed by a surveillance camera. Algorithms that focus on shadow elimination are mainly based on the morphological reconstruction which uses the surface spectral reflectance properties. Certainly, these latter properties are invariant to changes in lighting, scene composition, and geometry. Mathematical morphology has been constantly used in a variety of sensible issues related to image segmentation, noise filtering, shadow detection and elimination, etc. [26–28]. In fact, the mathematical morphology has a specific characteristic since it deals with discrete data in a nonlinear approach.
Logical dual concepts based on mathematical morphology in stratified institutions: applications to spatial reasoning
Published in Journal of Applied Non-Classical Logics, 2019
In mathematical morphology, erosion and dilation are operations that are defined, in a general deterministic and algebraic setting, on lattices, for instance on sets. Thus, they can be applied to formulas by identifying formulas with sets. We have two ways of doing this, either given a model M identifying a formula ϕ by the set of states η that satisfy ϕ, and classically denoted by , or identifying ϕ by the set of models that satisfy it. As usual in logic, our abstract dual operators based on morphological erosion and dilation will be studied both on sets of states and sets of models. The problem is that institutions do not explicitly provide, given a model M, its set of states. This is why we will define our abstract logical dual operators based on erosion and dilation in an extension of institutions, the stratified institutions (Aiguier & Diaconescu, 2007). Stratified institutions have been defined in Aiguier and Diaconescu (2007) as an extension of institutions to take into account the notion of open sentences, whose satisfaction is parametrised by sets of states. For instance, in first-order logic, the satisfaction is parametrised by the valuation of unbound variables, while in modal logics it is further parametrised by possible worlds. Hence, stratified institutions allow for a uniform treatment of such parametrizations of the satisfaction relation within the abstract setting of logics as institutions.