Advanced Comparative Analysis of the Hough Transform and RANSAC with Applications in Automated Growth Ring Detection in Otoliths and Integration with Deep Neural Network

Abstract

This study conducts a comprehensive comparative analysis of the Hough Transform and the RANSAC algorithm for the detection of growth rings in otolith images, which are critical indicators of fish population dynamics. Both techniques are assessed with respect to their robustness, detection accuracy, and resilience to noise when applied to complex and heterogeneous image datasets. To overcome the inherent limitations of these classical methods particularly in managing irregular, fragmented, or partially occluded patterns their integration with deep neural networks is further explored. The experimental results demonstrate that the fusion of conventional algorithms with deep learning architectures substantially improves both detection precision and the degree of automation. The proposed hybrid framework exhibits strong potential for the analysis of intricate biological structures and broader applications within the field of computer vision.