、import cv2
import numpy as np
def cv_show(img):
img=cv2.pyrDown((img))
cv2.imshow('name',img)
cv2.waitKey()
cv2.destroyAllWindows()
img1=cv2.imread('right_01.png')
img2=cv2.imread('left_01.png')
#建立生成器
sift=cv2.xfeatures2d.SIFT_create()
# 检测SIFT特征点,并计算描述子
kp1,des1=sift.detectAndCompute(img1,None)
kp2,des2=sift.detectAndCompute(img2,None)
kpp1=kp1.copy()
kpp2=kp2.copy()
kp1=np.float32([kp.pt for kp in kp1])
kp2=np.float32([kp.pt for kp in kp2])
#匹配两张图,所有特征
bf=cv2.BFMatcher()
matchers=bf.knnMatch(des1,des2,2)
good=[]
for m in matchers:
# 当最近距离跟次近距离的比值小于0.75值时,保留此匹配对
if len(m)==2 and m[0].distance<m[1].distance*0.75:
good.append((m[0].trainIdx,m[0].queryIdx))
#当筛选后的匹配对大于4,计算视角变换矩阵
if(len(good)>4):
pts1=np.float32([kp1[i] for (_,i) in good])
pts2=np.float32([kp2[i] for (i,_) in good])
#计算视角变换矩阵
(H,status) =cv2.findHomography(pts1,pts2,cv2.RANSAC,4.0)
#提取匹配结果
#H为3x3的视角变换矩阵
result=cv2.warpPerspective(img1,H,(img1.shape[1]+img2.shape[1],img1.shape[0]))
cv_show(result)
#将宁一张图片传入左边
result[0:img2.shape[0],0:img2.shape[1]]=img2.copy()
cv_show(result)
# print(status)
# w=cv2.drawMatches(img1,kpp1,img2,kpp1,good,flags=2)
# cv_show(w)
使用的图片: