本篇笔记主要记录Opencv里的图像翻转,平移,旋转,仿射及透视功能,主要是下面几个API:
cv2.flip() # 图像翻转
cv2.warpAffine() #图像仿射
cv2.getRotationMatrix2D() #取得旋转角度的Matrix
cv2.GetAffineTransform(src, dst, mapMatrix) #取得图像仿射的matrix
cv2.getPerspectiveTransform(src, dst) #取得图像透视的4个点起止值
cv2.warpPerspective() #图像透视
图像翻转 cv2.flip()
cv2.flip(src, flipCode[, dst]) → dst
src: 原始图像矩阵;
dst: 变换后的矩阵;
flipMode: 翻转模式,有三种模式
0 --- 垂直方向翻转; 1----- 水平方向翻转; -1:水平、垂直方向同时翻转
flipCode==0垂直翻转(沿X轴翻转),flipCode>0水平翻转(沿Y轴翻转),flipCode<0水平垂直翻转(先沿X轴翻转,再沿Y轴翻转,等价于旋转180°)
%matplotlib inline
from matplotlib import pyplot as plt
import numpy as np
import cv2
image = cv2.imread("aier.jpg")
# Flipped Horizontally 水平翻转
h_flip = cv2.flip(image, 1)
# Flipped Vertically 垂直翻转
v_flip = cv2.flip(image, 0)
# Flipped Horizontally & Vertically 水平垂直翻转
hv_flip = cv2.flip(image, -1)
plt.figure(figsize=(8,8))
plt.subplot(221)
plt.imshow(image[:,:,::-1])
plt.title('original')
plt.subplot(222)
plt.imshow(h_flip[:,:,::-1])
plt.title('horizontal flip')
plt.subplot(223)
plt.imshow(v_flip[:,:,::-1])
plt.title(' vertical flip')
plt.subplot(224)
plt.imshow(hv_flip[:,:,::-1])
plt.title('h_v flip')
# 调整子图间距
# plt.subplots_adjust(wspace=0.5, hspace=0.1)
plt.subplots_adjust(top=0.8, bottom=0.08, left=0.10, right=0.95, hspace=0,
wspace=0.35)
# plt.tight_layout()
plt.show()
cv2.flip()
图像平移,旋转,仿射 cv2.warpAffine()
cv2.warpAffine(src, M, dsize[, dst[, flags[, borderMode[, borderValue]]]])
src input image.
dst output image that has the size dsize and the same type as src .
M 2×3 transformation matrix.
dsize size of the output image.
flags combination of interpolation methods (see InterpolationFlags) and the optional flag WARP_INVERSE_MAP that means that M is the inverse transformation ( dst→src ).
borderMode pixel extrapolation method (see BorderTypes); when borderMode=BORDER_TRANSPARENT, it means that the pixels in the destination image corresponding to the "outliers" in the source image are not modified by the function.
borderValue value used in case of a constant border; by default, it is 0.
cv2.getRotationMatrix2D(center, angle, scale)
函数有三个参数:
center:图片的旋转中心
angle:旋转角度
scale:缩放比例,该例中0.5表示我们缩小一半
%matplotlib inline
from matplotlib import pyplot as plt
import cv2
import numpy as np
img = cv2.imread('aier.jpg')
rows,cols = img.shape[:2]
# 定义平移矩阵,需要是numpy的float32类型
# x轴平移200,y轴平移100, 2*3矩阵
M = np.float32([[1, 0, 200], [0, 1, 100]])
# 用仿射变换实现平移
img_s = cv2.warpAffine(img, M, (cols, rows), borderValue=(155, 150, 200))
# 第一个参数旋转中心,第二个参数旋转角度,第三个参数:缩放比例, 生成一2*3的矩阵
M = cv2.getRotationMatrix2D((cols/2,rows/2),90,1)
M1 = cv2.getRotationMatrix2D((cols/2,rows/2),180,1)
M2 = cv2.getRotationMatrix2D((cols/2,rows/2),60,1)
print(M)
'''
[[ 6.123234e-17 1.000000e+00 1.500000e+02]
[-1.000000e+00 6.123234e-17 6.500000e+02]]
'''
# 第三个参数:变换后的图像大小
img_tra = cv2.warpAffine(img,M,(cols,rows))
img_tra1 = cv2.warpAffine(img,M1,(cols,rows))
img_tra2 = cv2.warpAffine(img,M2,(cols,rows), borderValue=(155, 100, 155))
plt.figure(figsize=(8,8))
plt.subplot(221)
plt.imshow(img[:,:,::-1])
plt.subplot(222)
plt.imshow(img_s[:,:,::-1])
plt.subplot(223)
plt.imshow(img_tra[:,:,::-1])
plt.subplot(224)
plt.imshow(img_tra2[:,:,::-1])
plt.subplots_adjust(top=0.8, bottom=0.08, left=0.10, right=0.95, hspace=0,
wspace=0.35)
plt.show()
平移,旋转
图像仿射
图像的旋转加上拉升就是图像仿射变换,仿射变化也是需要一个M矩阵就可以,但是由于仿射变换比较复杂,一般直接找很难找到这个矩阵,opencv提供了根据变换前后三个点的对应关系来自动求解M。这个函数是
M=cv2.getAffineTransform(pos1,pos2),其中两个位置就是变换前后的对应位置关系。输出的就是仿射矩阵M。然后在使用函数cv2.warpAffine()。
cv.GetAffineTransform(src, dst, mapMatrix) → None
Parameters: 变换前的三个点与其对应的变换后的点.
src – Coordinates of triangle vertices in the source image.
dst – Coordinates of the corresponding triangle vertices in the destination image.
The function calculates the 2*3 matrix of an affine transform.
AffineMatrix = cv2.getAffineTransform(np.array(SrcPointsA),
np.array(CanvasPointsA))
图像仿射示例图
%matplotlib inline
from matplotlib import pyplot as plt
import cv2
import numpy as np
img = cv2.imread('aier.jpg')
rows,cols = img.shape[:2]
pts1 = np.float32([[50,50],[200,50],[50,200]])
pts2 = np.float32([[10,100],[200,20],[100,250]])
M = cv2.getAffineTransform(pts1,pts2)
#第三个参数:变换后的图像大小
res = cv2.warpAffine(img,M,(rows,cols))
plt.subplot(121)
plt.imshow(img[:,:,::-1])
plt.subplot(122)
plt.imshow(res[:,:,::-1])
plt.show()
透视 Perspective
视角变换,需要一个3*3变换矩阵。在变换前后要保证直线还是直线。
构建此矩阵需要在输入图像中找寻 4个点,以及在输出图像中对应的位置。这四个点中的任意三个点不能共线。
## pts1 ==> pts2
pts1=np.float32([[56,65],[368,52],[28,387],[389,390]])
pts2=np.float32([[0,0],[300,0],[0,300],[300,300]])
M=cv2.getPerspectiveTransform(pts1,pts2)
cv2.getPerspectiveTransform(np.array(SrcPointsA), np.array(CanvasPointsA))
cv2.getPerspectiveTransform(src, dst) → retval
cv2.warpPerspective(Img, PerspectiveMatrix, (300, 300))
cv2.warpPerspective(src, M, dsize[, dst[, flags[, borderMode[, borderValue]]]]) → dst
src – input image.
dst – output image that has the size dsize and the same type as src .
M – 3*3 transformation matrix.
dsize – size of the output image.
flags – combination of interpolation methods (INTER_LINEAR or INTER_NEAREST) and the optional flag WARP_INVERSE_MAP, that sets M as the inverse transformation ( dst ---> src ).
borderMode – pixel extrapolation method (BORDER_CONSTANT or BORDER_REPLICATE).
borderValue – value used in case of a constant border; by default, it equals 0.
4个点前后映射示例图
%matplotlib inline
from matplotlib import pyplot as plt
import cv2
import numpy as np
img=cv2.imread('aier.jpg')
rows,cols,ch=img.shape
pts1=np.float32([[56,5],[368,5],[28,387],[389,390]])
pts2=np.float32([[0,0],[300,0],[0,300],[300,300]])
M=cv2.getPerspectiveTransform(pts1,pts2)
print(M)
dst=cv2.warpPerspective(img,M,(300,300))
plt.subplot(121),plt.imshow(img),plt.title('Input')
plt.subplot(122),plt.imshow(dst),plt.title('Output')
plt.show()
图像透视
本文主要内容就是这些,其他更深入功能待后续继续完善.....
[参考]
Geometric Transformations of Images