ℱ‐homogeneous functions and a generalization of directional monotonicity. Issue 9 (17th January 2022)

• Record Type: Journal Article
• Authors: Santiago, Regivan; Sesma‐Sara, Mikel; Fernandez, Javier; Takac, Zdenko; Mesiar, Radko; Bustince, Humberto
• Is Part Of: International journal of intelligent systems. Volume 37:Issue 9(2022)
• Page Start: 5949
• Page End: 5970
• Publication Date: 2022-01-17
• Subjects: abstract homogeneity -- F‐homogeneity -- homogeneity; Artificial intelligence -- Periodicals; Expert systems (Computer science) -- Periodicals; Intelligence artificielle -- Périodiques; Systèmes experts (Informatique) -- Périodiques; 006.3
• DOI: 10.1002/int.22823
• Languages: English
• ISSNs: 0884-8173

Abstract

Abstract: A function that takes n numbers as input and outputs one number is said to be homogeneous whenever the result of multiplying each input by a certain factor λ yields the original output multiplied by that same factor. This concept has been extended by the notion of abstract homogeneity, which generalizes the product in the expression of homogeneity by a general function g and the effect of the factor λ by an automorphism. However, the effect of parameter λ remains unchanged for all the input values. In this study, we generalize further the condition of abstract homogeneity by introducing ℱ ‐homogeneity, which is defined with respect to a family of functions, enabling a different behavior for each of the inputs. Next, we study the properties that are satisfied by this family of functions and, moreover, we link this concept with the condition of directional monotonicity, which is a trendy property in the framework of aggregation functions. To achieve that, we generalize directional monotonicity by ℱ directional monotonicity, which is defined with respect to a family of functions ℱ and a family of vectors V . Finally, we show how the introduced concepts could be applied in two different problems of computer vision: a snow detection problem and image thresholding improvement.