Skip to main content
. 2022 Mar 23;22(7):2463. doi: 10.3390/s22072463

Figure 12.

Figure 12

Qualitative comparison between the proposed method and Deeplabv3+: (a) original RGB images; (b) ground truths; (c) results of Deeplabv3+; (d) results of the proposed method. The white region indicates the pixels that are correctly detected as FOD, the black indicates the pixels that are correctly detected as background, the yellow indicates the pixels that are incorrectly detected as FOD, and the green indicates the pixels that are incorrectly detected as background.