Skip to main content
. 2019 Jun 6;19(11):2577. doi: 10.3390/s19112577

Figure 8.

Figure 8

Segmentation of the four-class high-definition dataset (ac): test images, (df): segmented images using DeconvNet, (gi): segmented images using proposed light-weight network. Note that the color code for classes is: green: ground, red: vegetation, yellow: tree, blue: sky. Best viewed in color.