A Novel Edge Detection Operator Induced by ZA-Algebraic Filter Kernels with Applications in Digital Image Analysis

Abstract algebraic structures provide a rigorous mathematical framework for designing sophisticated, non-linear computational operators in digital image processing. In this paper, we introduce and formalize the theoretical foundations of ZA-algebras, a new class of algebraic systems governed by specific operational axioms and binary relations. In addition to the fundamentals of every ZA-algebra, its substructures and homomorphisms, we use the special structure of ZA-algebras to define an algebraic edge detection operator induced by ZA-algebra. In particular, an approach to dimension reduction is constructed to map 4 × 4 Cayley-based structural matrices to effective 3 × 3 filtering kernels (Mfinal) by localized algebraic additions, as well as a discrete Laplace operator and the weighting in the spatial domain. The proposed operator dynamically preserves high-frequency gradient boundaries while suppressing noise artifacts. The proposed ZA-filter was quantitatively evaluated against nine classical edge detection operators (Sobel, Prewitt, Canny, Roberts, Robinson, Kirsch, Marr-Hildreth, Standard Laplacian, and LoG) under the reported experimental conditions. The obtained results characterize the performance of the proposed approach across the evaluated images. In addition, selected recent frameworks (Fuzzy Logic, Fractional-Order derivative operators, and GLF-ResFormer) are discussed through descriptive comparisons based on published results, without implying a controlled head-to-head evaluation. Specifically, on benchmark datasets, the proposed method achieves Peak Signal-to-Noise Ratio (PSNR up to 11.5859 dB), Structural Similarity Index (SSIM up to 0.6137), Information Entropy of the pre-threshold filter responses (H = 6.1688 to 6.4685), Edge Pixel Ratio (EPR up to 0.1429), and boundary metrics on BSDS500 (ODS F-measure = 0.3331, OIS F-measure = 0.4803), with an average execution runtime of 0.001815 to 0.002710. Furthermore, non-parametric Wilcoxon signed-rank tests were employed to examine paired differences in absolute reconstruction error between the proposed ZA-filter and conventional edge detection operators. The reported results provide supplementary statistical evidence concerning reconstruction-error behavior, rather than direct evidence of superior boundary localization accuracy.

Authors

Institutions

Publication Details

Journal
Computers
Published
2026-11-16
DOI
https://doi.org/10.3390/computers15110844
Primary Topic
Image and Object Detection Techniques
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

A Novel Edge Detection Operator Induced by ZA-Algebraic Filter Kernels with Applications in Digital Image Analysis

Alaa Abdulhasan, Maytham Al-Hasooni, Marwa Alhamdany, Hiba Kareem et al.
Computers
Image and Object Detection Techniques
article

A Novel Edge Detection Operator Induced by ZA-Algebraic Filter Kernels with Applications in Digital Image Analysis

Alaa Abdulhasan, Maytham Al-Hasooni, Marwa Alhamdany, Hiba Kareem, Zainab Mousa
article en

Abstract

Abstract algebraic structures provide a rigorous mathematical framework for designing sophisticated, non-linear computational operators in digital image processing. In this paper, we introduce and formalize the theoretical foundations of ZA-algebras, a new class of algebraic systems governed by specific operational axioms and binary relations. In addition to the fundamentals of every ZA-algebra, its substructures and homomorphisms, we use the special structure of ZA-algebras to define an algebraic edge detection operator induced by ZA-algebra. In particular, an approach to dimension reduction is constructed to map 4 × 4 Cayley-based structural matrices to effective 3 × 3 filtering kernels (Mfinal) by localized algebraic additions, as well as a discrete Laplace operator and the weighting in the spatial domain. The proposed operator dynamically preserves high-frequency gradient boundaries while suppressing noise artifacts. The proposed ZA-filter was quantitatively evaluated against nine classical edge detection operators (Sobel, Prewitt, Canny, Roberts, Robinson, Kirsch, Marr-Hildreth, Standard Laplacian, and LoG) under the reported experimental conditions. The obtained results characterize the performance of the proposed approach across the evaluated images. In addition, selected recent frameworks (Fuzzy Logic, Fractional-Order derivative operators, and GLF-ResFormer) are discussed through descriptive comparisons based on published results, without implying a controlled head-to-head evaluation. Specifically, on benchmark datasets, the proposed method achieves Peak Signal-to-Noise Ratio (PSNR up to 11.5859 dB), Structural Similarity Index (SSIM up to 0.6137), Information Entropy of the pre-threshold filter responses (H = 6.1688 to 6.4685), Edge Pixel Ratio (EPR up to 0.1429), and boundary metrics on BSDS500 (ODS F-measure = 0.3331, OIS F-measure = 0.4803), with an average execution runtime of 0.001815 to 0.002710. Furthermore, non-parametric Wilcoxon signed-rank tests were employed to examine paired differences in absolute reconstruction error between the proposed ZA-filter and conventional edge detection operators. The reported results provide supplementary statistical evidence concerning reconstruction-error behavior, rather than direct evidence of superior boundary localization accuracy.

ComputersVol. 15(11)
University of Kufa (IQ)
Image and Object Detection Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon