1+ import cv2
2+ import numpy as np
3+ import matplotlib .pyplot as plt
4+ import math
5+ 6+ # 读取图片
7+ img = cv2 .imread ('me1.jpg' )
8+ src = cv2 .cvtColor (img , cv2 .COLOR_BGR2RGB )
9+ 10+ # 新建目标图像
11+ dst1 = np .zeros_like (img )
12+ 13+ # 获取图像行和列
14+ rows , cols = img .shape [:2 ]
15+ 16+ # --------------毛玻璃效果--------------------
17+ # 像素点邻域内随机像素点的颜色替代当前像素点的颜色
18+ offsets = 5
19+ random_num = 0
20+ for y in range (rows - offsets ):
21+ for x in range (cols - offsets ):
22+ random_num = np .random .randint (0 , offsets )
23+ dst1 [y , x ] = src [y + random_num , x + random_num ]
24+ 25+ # -------油漆特效------------
26+ # 图像灰度处理
27+ gray = cv2 .cvtColor (img , cv2 .COLOR_BGR2GRAY )
28+ 29+ # 自定义卷积核
30+ kernel = np .array ([[- 1 , - 1 , - 1 ], [- 1 , 10 , - 1 ], [- 1 , - 1 , - 1 ]])
31+ 32+ # 图像浮雕效果
33+ dst2 = cv2 .filter2D (gray , - 1 , kernel )
34+ 35+ # ----------素描特效-------------
36+ # 高斯滤波降噪
37+ gaussian = cv2 .GaussianBlur (gray , (5 , 5 ), 0 )
38+ 39+ # Canny算子
40+ canny = cv2 .Canny (gaussian , 50 , 150 )
41+ 42+ # 阈值化处理
43+ ret , dst3 = cv2 .threshold (canny , 100 , 255 , cv2 .THRESH_BINARY_INV )
44+ 45+ # -------怀旧特效-----------------
46+ # 新建目标图像
47+ dst4 = np .zeros ((rows , cols , 3 ), dtype = "uint8" )
48+ 49+ # 图像怀旧特效
50+ for i in range (rows ):
51+ for j in range (cols ):
52+ B = 0.272 * img [i , j ][2 ] + 0.534 * img [i , j ][1 ] + 0.131 * img [i , j ][0 ]
53+ G = 0.349 * img [i , j ][2 ] + 0.686 * img [i , j ][1 ] + 0.168 * img [i , j ][0 ]
54+ R = 0.393 * img [i , j ][2 ] + 0.769 * img [i , j ][1 ] + 0.189 * img [i , j ][0 ]
55+ if B > 255 :
56+ B = 255
57+ if G > 255 :
58+ G = 255
59+ if R > 255 :
60+ R = 255
61+ dst4 [i , j ] = np .uint8 ((B , G , R ))
62+ 63+ # ---------------光照特效--------------------
64+ # 设置中心点
65+ centerX = rows / 2
66+ centerY = cols / 2
67+ print (centerX , centerY )
68+ radius = min (centerX , centerY )
69+ print (radius )
70+ 71+ # 设置光照强度
72+ strength = 200
73+ 74+ # 新建目标图像
75+ dst5 = np .zeros ((rows , cols , 3 ), dtype = "uint8" )
76+ 77+ # 图像光照特效
78+ for i in range (rows ):
79+ for j in range (cols ):
80+ # 计算当前点到光照中心的距离(平面坐标系中两点之间的距离)
81+ distance = math .pow ((centerY - j ), 2 ) + math .pow ((centerX - i ), 2 )
82+ # 获取原始图像
83+ B = src [i , j ][0 ]
84+ G = src [i , j ][1 ]
85+ R = src [i , j ][2 ]
86+ if (distance < radius * radius ):
87+ # 按照距离大小计算增强的光照值
88+ result = (int )(strength * (1.0 - math .sqrt (distance ) / radius ))
89+ B = src [i , j ][0 ] + result
90+ G = src [i , j ][1 ] + result
91+ R = src [i , j ][2 ] + result
92+ # 判断边界 防止越界
93+ B = min (255 , max (0 , B ))
94+ G = min (255 , max (0 , G ))
95+ R = min (255 , max (0 , R ))
96+ dst5 [i , j ] = np .uint8 ((B , G , R ))
97+ else :
98+ dst5 [i , j ] = np .uint8 ((B , G , R ))
99+ 100+ # --------------怀旧特效-----------------
101+ # 新建目标图像
102+ dst6 = np .zeros ((rows , cols , 3 ), dtype = "uint8" )
103+ 104+ # 图像流年特效
105+ for i in range (rows ):
106+ for j in range (cols ):
107+ # B通道的数值开平方乘以参数12
108+ B = math .sqrt (src [i , j ][0 ]) * 12
109+ G = src [i , j ][1 ]
110+ R = src [i , j ][2 ]
111+ if B > 255 :
112+ B = 255
113+ dst6 [i , j ] = np .uint8 ((B , G , R ))
114+ 115+ # ------------卡通特效-------------------
116+ # 定义双边滤波的数目
117+ num_bilateral = 7
118+ 119+ # 用高斯金字塔降低取样
120+ img_color = src
121+ 122+ # 双边滤波处理
123+ for i in range (num_bilateral ):
124+ img_color = cv2 .bilateralFilter (img_color , d = 9 , sigmaColor = 9 , sigmaSpace = 7 )
125+ 126+ # 灰度图像转换
127+ img_gray = cv2 .cvtColor (src , cv2 .COLOR_RGB2GRAY )
128+ 129+ # 中值滤波处理
130+ img_blur = cv2 .medianBlur (img_gray , 7 )
131+ 132+ # 边缘检测及自适应阈值化处理
133+ img_edge = cv2 .adaptiveThreshold (img_blur , 255 ,
134+ cv2 .ADAPTIVE_THRESH_MEAN_C ,
135+ cv2 .THRESH_BINARY ,
136+ blockSize = 9 ,
137+ C = 2 )
138+ 139+ # 转换回彩色图像
140+ img_edge = cv2 .cvtColor (img_edge , cv2 .COLOR_GRAY2RGB )
141+ 142+ # 与运算
143+ dst6 = cv2 .bitwise_and (img_color , img_edge )
144+ 145+ # ------------------均衡化特效--------------------
146+ # 新建目标图像
147+ dst7 = np .zeros ((rows , cols , 3 ), dtype = "uint8" )
148+ 149+ # 提取三个颜色通道
150+ (b , g , r ) = cv2 .split (src )
151+ 152+ # 彩色图像均衡化
153+ bH = cv2 .equalizeHist (b )
154+ gH = cv2 .equalizeHist (g )
155+ rH = cv2 .equalizeHist (r )
156+ 157+ # 合并通道
158+ dst7 = cv2 .merge ((bH , gH , rH ))
159+ 160+ # -----------边缘特效---------------------
161+ # 高斯滤波降噪
162+ gaussian = cv2 .GaussianBlur (gray , (3 , 3 ), 0 )
163+ 164+ # Canny算子
165+ # dst8 = cv2.Canny(gaussian, 50, 150)
166+ 167+ # Scharr算子
168+ x = cv2 .Scharr (gaussian , cv2 .CV_32F , 1 , 0 ) # X方向
169+ y = cv2 .Scharr (gaussian , cv2 .CV_32F , 0 , 1 ) # Y方向
170+ absX = cv2 .convertScaleAbs (x )
171+ absY = cv2 .convertScaleAbs (y )
172+ dst8 = cv2 .addWeighted (absX , 0.5 , absY , 0.5 , 0 )
173+ 174+ 175+ # 用来正常显示中文标签
176+ plt .rcParams ['font.sans-serif' ] = ['SimHei' ]
177+ 178+ # 循环显示图形
179+ titles = ['原图' , '毛玻璃' , '浮雕' , '素描' , '怀旧' , '光照' , '卡通' , '均衡化' , '边缘' ]
180+ images = [src , dst1 , dst2 , dst3 , dst4 , dst5 , dst6 , dst7 , dst8 ]
181+ for i in range (9 ):
182+ plt .subplot (3 , 3 , i + 1 ), plt .imshow (images [i ], 'gray' )
183+ plt .title (titles [i ])
184+ plt .xticks ([]), plt .yticks ([])
185+ 186+ if __name__ == '__main__' :
187+ 188+ plt .show ()
0 commit comments