I need to do a CODE for the first stage of my project which is using remotely sensed images for automatic classification of urban materials (WITHOUT using any commercial software such as arcGIS, ...) to distinguish land-cover objects (e.g. buildings, parking lots, roads, soil, grass, and water basins). The idea is using more than one classification method (such as Object-based Image Analysis, Knowledge-based method, Convolutional Neural Network, ...). Then make a comparison among them (as a table) to find which one is more accurecy. This is the first stage of my project which I need it it for now.