École Nationale Supérieure de
Techniques Avancées

Abstract

The watershed algorithm is a segmentation technique widely used in various fields, such as medical imaging and computer vision. Inspired by hydrographic basins, the image is divided into distinct regions based on the intensity of the pixels, treating it as a three-dimensional topography. Interactive watershed is a variation of the traditional method that incorporates user interaction into the segmentation process with the aim of combining the robustness of the traditional method with the precision of human intervention, making it a powerful tool for object segmentation. This project presents the theoretical background of different interactive watershed algorithms and its implementation details. A comparative analysis is conducted to assess the functionality and applicability of these methods. Furthermore, a study of the execution times is presented, examining how the algorithms perform across multiple user interactions and varying data sizes.