Python implements image stitching

The examples in this article share the specific code of python to realize image stitching for your reference.

1. Image to be stitched

2. Image feature point matching results based on SIFT feature points and RANSAC method

3. Image transformation result

4. Code and precautions

import cv2
import numpy as np
 
 
def cv_show(name, image):
 cv2.imshow(name, image)
 cv2.waitKey(0)
 cv2.destroyAllWindows()
 
 
def detectAndCompute(image):
 image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
 sift = cv2.xfeatures2d.SIFT_create()(kps, features)= sift.detectAndCompute(image, None)
 kps = np.float32([kp.pt for kp in kps]) #The points obtained need to be further converted before they can be used
 return(kps, features)
 
 
def matchKeyPoints(kpsA, kpsB, featuresA, featuresB, ratio =0.75, reprojThresh =4.0):
 # ratio is the recommended threshold for nearest neighbor matching
 # reprojThresh is the recommended threshold for random sampling consistency
 matcher = cv2.BFMatcher()
 rawMatches = matcher.knnMatch(featuresA, featuresB,2)
 matches =[]for m in rawMatches:iflen(m)==2 and m[0].distance < ratio * m[1].distance:
 matches.append((m[0].queryIdx, m[0].trainIdx))
 kpsA = np.float32([kpsA[m[0]]for m in matches]) #Use np.float32 conversion list
 kpsB = np.float32([kpsB[m[1]]for m in matches])(M, status)= cv2.findHomography(kpsA, kpsB, cv2.RANSAC, reprojThresh)return(M, matches, status) #Not all points have matching solutions, their status is stored in the status
 
 
def stich(imgA, imgB, M):
 result = cv2.warpPerspective(imgA, M,(imgA.shape[1]+ imgB.shape[1], imgA.shape[0]))
 result[0:imageA.shape[0],0:imageB.shape[1]]= imageB
 cv_show('result', result)
 
 
def drawMatches(imgA, imgB, kpsA, kpsB, matches, status):(hA, wA)= imgA.shape[0:2](hB, wB)= imgB.shape[0:2]
 # Note the 3 channels and uint8 type here
 drawImg = np.zeros((max(hA, hB), wA + wB,3),'uint8')
 drawImg[0:hB,0:wB]= imageB
 drawImg[0:hA, wB:]= imageA
 for((queryIdx, trainIdx),s)inzip(matches, status):if s ==1:
 # Note that float32--  int
 pt1 =(int(kpsB[trainIdx][0]),int(kpsB[trainIdx][1]))
 pt2 =(int(kpsA[trainIdx][0])+ wB,int(kpsA[trainIdx][1]))
 cv2.line(drawImg, pt1, pt2,(0,0,255))cv_show("drawImg", drawImg)
 
 
# Read image
imageA = cv2.imread('./right_01.png')cv_show("imageA", imageA)
imageB = cv2.imread('./left_01.png')cv_show("imageB", imageB)
# Calculate SIFT feature points and feature vectors(kpsA, featuresA)=detectAndCompute(imageA)(kpsB, featuresB)=detectAndCompute(imageB)
# Obtain a homography matrix based on nearest neighbor and random sampling consistency(M, matches, status)=matchKeyPoints(kpsA, kpsB, featuresA, featuresB)
# Draw matching results
drawMatches(imageA, imageB, kpsA, kpsB, matches, status)
# Splicing
stich(imageA, imageB, M)

The above is the whole content of this article, I hope it will be helpful to everyone's study.