Table 3. Parameters of the proposed patch-wise U-net model.

No	Layer name	Type	Output Shape	No. of Parameters	Connected to
1	input_1	Image	128×128×1	0	-
2	conv2d_1	2D Convolution	128×128×64	640	input_1
3	conv2d_2	2D Convolution	128×128×64	36928	conv2d_1
4	max_pooling2d_1	Max Pooling 2D	64×64×64	0	conv2d_2
5	conv2d_3	2D Convolution	64×64×128	73856	max_pooling2d_1
6	conv2d_4	2D Convolution	64×64×128	147584	conv2d_3
7	max_pooling2d_2	Max Pooling 2D	32×32×128	0	conv2d_4
8	conv2d_5	2D Convolution	32×32×256	295168	max_pooling2d_2
9	conv2d_6	2D Convolution	32×32×256	590080	conv2d_5
10	max_pooling2d_3	Max Pooling 2D	16×16×256	0	conv2d_6
11	conv2d_7	2D Convolution	16×16×512	1180160	max_pooling2d_3
12	conv2d_8	2D Convolution	16×16×512	2359808	conv2d_7
13	up_sampling2d_1	Up Sampling 2D	32×32×512	0	conv2d_8
14	conv2d_9	2D Convolution	32×32×256	524544	up_sampling2d_1
15	concatenated_1	Concatenate	32×32×512	0	conv2d_6
					conv2d_9
16	conv2d_10	2D Convolution	32×32×256	1179904	concatenated_1
17	conv2d_11	2D Convolution	32×32×256	590080	conv2d_10
18	up_sampling2d_2	Up Sampling 2D	64×64×256	0	conv2d_11
19	conv2d_12	2D Convolution	64×64×128	131200	up_sampling2d_2
20	concatenated_2	Concatenate	64×64×256	0	conv2d_4
					conv2d_12
21	conv2d_13	2D Convolution	64×64×128	295040	concatenated_2
22	conv2d_14	2D Convolution	64×64×128	147584	conv2d_13
23	up_sampling2d_3	Up Sampling 2D	128×128×128	0	conv2d_14
24	conv2d_15	2D Convolution	128×128×64	32832	up_sampling2d_3
25	concatenated_3	Concatenate	128×128×128	0	conv2d_2
					conv2d_15
26	conv2d_16	2D Convolution	128×128×64	73792	concatenated_3
27	conv2d_17	2D Convolution	128×128×64	36928	conv2d_16
28	conv2d_18	2D Convolution	128×128×4	260	conv2d_17