本文按照如下设计
ImageStitching_ExcessThree.py
from Stitcher import Stitcher
import cv2
import my_utils
# 只拼接两张图片# 读取需要拼接的图片
# imageA_original = cv2.imread("left_01.png")
# imageB_original = cv2.imread("right_01.png")
imageA_original = cv2.imread("left_01.jpg")
imageB_original = cv2.imread("right_01.jpg")
imageC_original = cv2.imread("right_02.jpg")# 图像预处理-改变图像大小
imageA = my_utils.resize(imageA_original,width=500)
imageB = my_utils.resize(imageB_original,width=500)
imageC = my_utils.resize(imageC_original,width=500)# 把图片拼接成全景图
stitcher = Stitcher()
(result, vis) = stitcher.stitch([imageA, imageB], showMatches=True)
(result, vis) = stitcher.stitch([result, imageC], showMatches=True)# 显示所有图片
stitcher.cv_show("Image A", imageA)
stitcher.cv_show("Image B", imageB)
stitcher.cv_show("Image C", imageC)
stitcher.cv_show("Keypoint Matches", vis)
stitcher.cv_show("Result", result)ImageStitching_JustTwo.py
from Stitcher import Stitcher
import cv2
import my_utils
# 只拼接两张图片# 读取需要拼接的图片
# imageA_original = cv2.imread("left_01.png")
# imageB_original = cv2.imread("right_01.png")
imageA_original = cv2.imread("left_01.jpg")
imageB_original = cv2.imread("right_01.jpg")# 图像预处理-改变图像大小
imageA = my_utils.resize(imageA_original,width=500)
imageB = my_utils.resize(imageB_original,width=500)# 把图片拼接成全景图
stitcher = Stitcher()
(result, vis) = stitcher.stitch([imageA, imageB], showMatches=True)
print(vis)
# 显示所有图片
stitcher.cv_show("Image A", imageA)
stitcher.cv_show("Image B", imageB)
stitcher.cv_show("Keypoint Matches", vis)
stitcher.cv_show("Result", result)
my_utils.py
import cv2def resize(image, width=None, height=None, inter=cv2.INTER_AREA):dim = None(h, w) = image.shape[:2]if width is None and height is None:return imageif width is None:r = height / float(h)dim = (int(w * r), height)print('width is None', dim)else:r = width / float(w)dim = (width, int(h * r))print('height is None', dim)resized = cv2.resize(image, dim, interpolation=inter)return resized
Stitcher.py
import numpy as np
import cv2class Stitcher:# 拼接函数def stitch(self, images, ratio=0.75, reprojThresh=4.0, showMatches=False):# 获取图片(imageB, imageA) = images# 检测A,B图片的SIFT关键特征点，并且计算其特征描述子(特征向量)(kpsA, featureA) = self.detectAndDescribe(imageA)(kpsB, featureB) = self.detectAndDescribe(imageB)# 匹配两张图片的所有特征点，返回匹配结果M = self.matchKeypoints(kpsA, kpsB, featureA, featureB, ratio, reprojThresh)# 如果返回结果为空，没有匹配成功的特征点，退出算法if M is None:return None# 否则，提取匹配结果# H是3x3视角变换矩阵(matches, H, status) = M# 将图片A进行视角变换，result是变换后图片result = cv2.warpPerspective(imageA, H, (imageA.shape[1] + imageB.shape[1], imageA.shape[0]))# self.cv_show('result', result)# 将图片B传入result图片最左端result[0:imageB.shape[0], 0:imageB.shape[1]] = imageB# self.cv_show('result', result)# 检测是否需要显示图片匹配if showMatches:# 生成匹配图片vis = self.drawMatches(imageA, imageB, kpsA, kpsB, matches, status)# 返回结果return (result, vis)# 返回匹配结果return result# 获取图片关键点和特征描述子def detectAndDescribe(self, img):# 转灰度图gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)# 建立SIFT实例对象descriptor = cv2.xfeatures2d.SIFT_create()# 检测SIFT特征点并且计算描述子(kps, feature) = descriptor.detectAndCompute(img, None)# 并且将结果转换为Numpy数组kps = np.float32([kp.pt for kp in kps])# 返回特征点集合描述子return (kps, feature)# 匹配两张图片的特征点def matchKeypoints(self, kpsA, kpsB, featureA, featureB, ratio, reprojThresh):# 使用BF匹配matcher = cv2.BFMatcher()# 使用KNN来对A,B图的SIFT特征进行匹配rawMatches = matcher.knnMatch(featureA, featureB, k=2)# 获得较好的相匹配的特征（筛选特征点）matches = []for m in rawMatches:# 当最近距离跟次近距离的比值小于ratio值时，保留此匹配对if len(m) == 2 and m[0].distance < m[1].distance * ratio:# 存储两个点在featuresA, featuresB中的索引值# DMatch.trainIdx - Index of the descriptor in train descriptors# DMatch.queryIdx - Index of the descriptor in query descriptorsmatches.append((m[0].trainIdx, m[0].queryIdx))# 当筛选后的匹配对大于4时，计算视角变换矩阵if len(matches) > 4:# 获取匹配对的点的坐标ptsA = np.float32([kpsA[i] for (_, i) in matches])ptsB = np.float32([kpsB[i] for (i, _) in matches])# 计算视角变换矩阵(H, status) = cv2.findHomography(ptsA, ptsB, cv2.RANSAC, reprojThresh)# 返回匹配结果和单应性矩阵和statusreturn (matches, H, status)# 如果匹配对小于4时，返回Nonereturn None# 展示图片def cv_show(self, name, img):cv2.imshow(name, img)cv2.waitKey(0)cv2.destroyAllWindows()def drawMatches(self, imageA, imageB, kpsA, kpsB, matches, status):# 初始化可视化图片，将A、B图左右连接到一起print("Start draw Matching...")(hA, wA) = imageA.shape[:2](hB, wB) = imageB.shape[:2]vis = np.zeros((max(hA, hB), wA + wB, 3), dtype="uint8")vis[0:hA, 0:wA] = imageAvis[0:hB, wA:] = imageB# 联合遍历，画出匹配对for ((trainIdx, queryIdx), s) in zip(matches, status):# 当点对匹配成功时，画到可视化图上if s == 1:# 画出匹配对ptA = (int(kpsA[queryIdx][0]), int(kpsA[queryIdx][1]))ptB = (int(kpsB[trainIdx][0]) + wA, int(kpsB[trainIdx][1]))cv2.line(vis, ptA, ptB, (0, 255, 0), 1)# 返回可视化结果return vis