Architecture of Deep Convolutional Encoder-Decoder Networks for Building Footprint Semantic Segmentation

Figure 9

Examples of data augmentation pre-processing. (a) Transform image and (b) transform ground truth mask.

(a)

(b)