In this report the work that has been conducted within the R&D-project Plattformbundna UV-sensorsystem for automating the detection of mines with machine learning methods is presented. The framework used to achieve this automation is known as change detection (CD), which involves comparing images taken at different times to identify significant differences. Earlier a work has been carried out with emphasis on classical methods. These methods are described in greater detail in an earlier edition of the final report for the Plattformbundna uv-sensorsystem project (FOI-R--4697--SE). The data has been collected in collaboration with 33:rd and 42:nd Mine Countermeasure Divisions, and MWDC (Maritime Warfare Data centre), which are a part of the Swedish Armed Forces responsible for categorizing and analysing data from the Swedish underwater domain. This report provides an overview of change detection techniques, with focus on machine learning methods, and their application to mine detection.