Semantic Visual Segmentation of a Mobile Robot Environment Using Deep Learning Model