Research Article
Pavement Crack Detection and Segmentation Method Based on Improved Deep Learning Fusion Model
Table 3
U-Net convolution feature layers and feature map outputs.
| VGG16 layer (type) | Resnet layer (type) | Output shape |
| block1_conv2 | leaky_re_lu_1 | (512, 512, 64) | block2_conv2 | leaky_re_lu_4 | (256, 256, 128) | block3_conv3 | leaky_re_lu_9 | (128, 128, 256) | block4_conv3 | leaky_re_lu_26 | (64, 64, 512) | dropout_1 | leaky_re_lu_43 | (32, 32, 1024) |
|
|