Our project aims to develop a machine learning-based approach to image segmentation using multi-sensor imagery. By utilising a dataset that contains Multispectral and Panchromatic pair images, we plan to train a model that can accurately segment an image into different regions. The model will be evaluated on its performance using standard metrics such as accuracy, precision, and recall. We will also provide qualitative results to demonstrate the effectiveness of the proposed approach.
Image segmentation is a useful technique for analysing satellite imagery for a variety of reasons. Some specific examples of the uses of image segmentation in satellite imagery include; Land use and land cover mapping, Monitoring natural resources , Disaster response and recovery, Infrastructure monitoring , Military and security and many more. Image segmentation is a powerful technique for extracting information from satellite images and can be applied to a wide variety of applications, depending on the specific needs of the task
The use of machine learning in image segmentation is becoming highly favoured, as it allows us to automatically and accurately identify and locate objects or structures within an image. By training a model on a labelled dataset, it can learn to recognize patterns and features that correspond to different objects or regions within an image, and then use that knowledge to identify those objects or regions in new images.
Using multi-sensor imagery for image segmentation can improve the accuracy and robustness of the segmentation results. Each sensor captures images at different wavelengths or with different properties, such as colour, infrared, or thermal. These different types of information can provide complementary information about the objects or regions in an image, which can make it easier for the machine learning model to accurately identify and locate them.