-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathfunctions.py
120 lines (109 loc) · 5.2 KB
/
functions.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
#-*-coding:UTF-8 -*-
import numpy as np
import gradio as gr
import cv2 as cv
import math as math
from helper_functions import *
from DCGAN import *
def function_hw1(input_image: gr.Image, H, L, S):
if input_image is None:
raise gr.Error('输入错误：在处理之前请先输入一张图像', duration=5)
hls_image = cv.cvtColor(input_image, cv.COLOR_RGB2HLS)
hls_image = HSL_adjust(hls_image, H, L, S)
output_image = cv.cvtColor(hls_image, cv.COLOR_HLS2RGB)
my_hls_image = my_cvtColor_RGB2HLS(input_image)
# my_hls_image = cv.cvtColor(input_image, cv.COLOR_RGB2HLS)
my_hls_image = HSL_adjust(my_hls_image, H, L, S)
my_output_image = my_cvtColor_HLS2RGB(my_hls_image)
# my_output_image = cv.cvtColor(my_hls_image, cv.COLOR_HLS2RGB)
return output_image, my_output_image
def function_hw2(input_image, size_x, size_y, mode, rotate_angle, biascut_percent):
if input_image is None:
raise gr.Error('输入错误：在处理之前请先输入一张图像', duration=5)
match mode:
case '最近邻插值':
mode = cv.INTER_NEAREST
case '双线性插值':
mode = cv.INTER_LINEAR
case '双三次插值':
mode = cv.INTER_CUBIC
case 'Lanczos插值':
mode = cv.INTER_LANCZOS4
case _:
mode = cv.INTER_LINEAR
# resize
output_image = cv.resize(input_image, None, fx=size_x, fy=size_y, interpolation=mode)
output_image2 = my_resize(input_image, size_x, size_y, mode)
# rotate
match rotate_angle:
case '0':
pass
case '90':
output_image = cv.rotate(output_image, cv.ROTATE_90_CLOCKWISE)
output_image2 = cv.rotate(output_image2, cv.ROTATE_90_CLOCKWISE)
case '180':
output_image = cv.rotate(output_image, cv.ROTATE_180)
output_image2 = cv.rotate(output_image2, cv.ROTATE_180)
case '270':
output_image = cv.rotate(output_image, cv.ROTATE_90_COUNTERCLOCKWISE)
output_image2 = cv.rotate(output_image2, cv.ROTATE_90_COUNTERCLOCKWISE)
case _:
pass
# biascut
output_image = cv.warpAffine(output_image, np.float32([[1, biascut_percent, 0], [0, 1, 0]]), (output_image.shape[1], output_image.shape[0]))
output_image2 = my_warp_affine(output_image2, np.float32([[1, biascut_percent, 0], [0, 1, 0]]), (output_image2.shape[0], output_image2.shape[1]))
diff_image = my_diff_image(output_image, output_image2)
return output_image, output_image2, diff_image
def function_hw3(seed, gender):
if seed is None:
raise gr.Error('内部错误', duration=5)
if seed == 0:
seed = np.random.randint(0, 2147483647)
output_image = generate_image(seed, gender)
return output_image
def function_hw4(input_image, mode):
if input_image is None:
raise gr.Error('输入错误：在处理之前请先输入一张图像', duration=5)
match mode:
case '双边滤波':
output_image = cv.bilateralFilter(input_image, 5, 75, 75)
case 'NLM滤波':
output_image = cv.fastNlMeansDenoisingColored(input_image)
case '导向滤波':
output_image = cv.ximgproc.guidedFilter(input_image, input_image, 5, 0.1)
case '手动导向滤波':
output_image = my_guided_filter(input_image, input_image, 5, 0.1)
case '基于盒式滤波优化的快速导向滤波':
output_image = my_fast_guided_filter(input_image, input_image, 5, 0.1)
case '手动双边滤波':
output_image = my_bilateral_filter(input_image, 5, 75, 75)
case _:
output_image = input_image
output_image2 = my_diff_image(input_image, output_image)
output_image2 = cv.cvtColor(output_image2, cv.COLOR_RGB2GRAY)
return output_image, output_image2
def function_hw5(input_image, input_image2, mode):
if input_image is None:
raise gr.Error('输入错误：在处理之前请先输入一张图像', duration=5)
match mode:
case '直方图均衡化':
output_image = cv.cvtColor(input_image, cv.COLOR_RGB2HLS)
output_image[:, :, 1] = cv.equalizeHist(output_image[:, :, 1])
case 'CLAHE':
clahe = cv.createCLAHE()
output_image = cv.cvtColor(input_image, cv.COLOR_RGB2HLS)
output_image[:, :, 1] = clahe.apply(output_image[:, :, 1])
case '手动直方图均衡化':
output_image = cv.cvtColor(input_image, cv.COLOR_RGB2HLS)
output_image[:, :, 1] = my_equalizeHist(output_image[:, :, 1])
case '手动CLAHE':
clahe = my_createCLAHE()
output_image = cv.cvtColor(input_image, cv.COLOR_RGB2HLS)
output_image[:, :, 1] = clahe.apply(output_image[:, :, 1])
case '直方图匹配':
if input_image2 is None:
raise gr.Error('输入错误：在处理之前请先输入一张匹配图像', duration=5)
output_image = cv.cvtColor(input_image, cv.COLOR_RGB2HLS)
input_image2 = cv.cvtColor(input_image2, cv.COLOR_RGB2HLS)
output_image[:, :, 1] = my_hist_match(output_image[:, :, 1], input_image2[:, :, 1])
return cv.cvtColor(output_image, cv.COLOR_HLS2RGB)