The main objective of this study is to map and monitor slums using geoinformation technologies with more focus on the comparison of GIS and image analysis methods, of remotely sensed imagery (i.e. pixel-based, object-based and deep learning) and their algorithms (maximum likelihood, random trees, support vector machine and U-Net classifier) thereby choosing the optimal algorithm for slum mapping. Two study areas were chosen for this thesis: Lagos Mainland LGA (Nigeria) and Vila Andrade district (Brazil). The dataset used are Sentinel-2 imagery of the two study areas, drone imagery of Lagos Mainland LGA, orthophoto of Vila Andrade district and their respective administrative boundaries. This study adopted the different methods within the overall strategy supervised image classification where classification schema was created with five classes (slums, non-slums, vegetation, water and roads). Training samples were selected from each imagery which were then used to train the algorithms for the classification proper. The classification results for all dataset were assessed using the site-specific accuracy assessment including error matrix. The results were published as Web Map Service (classification results) and WFS (the administrative boundaries) using Geoserver to aid their usage in the web application environment. The web application was developed with a leaflet software and VS code editor to visualize the results. The results of this study showed that the SVM algorithm outperformed other algorithms within the pixel and object-based methods, although the object-based SVM performed better with an overall accuracy of 68% over the pixel-based SVM (63.1%). The RT algorithm for both pixel and object-based methods had 58.4% and 52.8% accuracy, followed by the ML algorithm with an overall accuracy of 49.8% and 38.7% for both methods. The deep learning U-Net algorithm had an overall accuracy of 60%.The main objective of this study is to map and monitor slums using geoinformation technologies with more focus on the comparison of GIS and image analysis methods, of remotely sensed imagery (i.e. pixel-based, object-based and deep learning) and their algorithms (maximum likelihood, random trees, support vector machine and U-Net classifier) thereby choosing the optimal algorithm for slum mapping. Two study areas were chosen for this thesis: Lagos Mainland LGA (Nigeria) and Vila Andrade district (Brazil). The dataset used are Sentinel-2 imagery of the two study areas, drone imagery of Lagos Mainland LGA, orthophoto of Vila Andrade district and their respective administrative boundaries. This study adopted the different methods within the overall strategy supervised image classification where classification schema was created with five classes (slums, non-slums, vegetation, water and roads). Training samples were selected from each imagery which were then used to train the algorithms for the classification proper. The classification results for all dataset were assessed using the site-specific accuracy assessment including error matrix. The results were published as Web Map Service (classification results) and WFS (the administrative boundaries) using Geoserver to aid their usage in the web application environment. The web application was developed with a leaflet software and VS code editor to visualize the results. The results of this study showed that the SVM algorithm outperformed other algorithms within the pixel and object-based methods, although the object-based SVM performed better with an overall accuracy of 68% over the pixel-based SVM (63.1%). The RT algorithm for both pixel and object-based methods had 58.4% and 52.8% accuracy, followed by the ML algorithm with an overall accuracy of 49.8% and 38.7% for both methods. The deep learning U-Net algorithm had an overall accuracy of 60%.