OpenCV对齐两张图片

最新推荐文章于 2024-07-01 14:04:41 发布

MiniSunFlower

最新推荐文章于 2024-07-01 14:04:41 发布

阅读量6.7k

点赞数 2

文章标签： OpenCv

本文链接：https://blog.csdn.net/yoozaishi/article/details/81000174

版权

比较两幅图像的异同时需要先对两幅图像进行对齐操作

对齐方法如下：

1 使用OpenCV对两幅图像进行角点检测

2 求取两幅图像的角点平均值

3 比较两幅图像的角点平均值获得偏移值

4 根据偏移值矫正图像

代码如下：

import numpy
import cv2
def transformation_from_points(cropImg1,cropImg2):
    corners1 = cv2.goodFeaturesToTrack(cropImg1,25,0.01,10)
    n = int(len(corners1))
    corners1 = corners1.reshape(n, 2)
    corners2 = cv2.goodFeaturesToTrack(cropImg2,25,0.01,10)
    n = int(len(corners2))
    corners2 = corners2.reshape(n, 2)

    points1 = corners1
    points2 = corners2
    sumX1 = 0
    sumY1 = 0
    x,y = points1.shape
    for m in range(x):
        sumX1 = sumX1 + points1[m][0]
        sumY1 = sumY1 + points2[m][1]
    sumX2 = 0
    sumY2 = 0
    x,y = points2.shape
    for n in range(x):
        sumX2 = sumX2 + points2[n][0]
        sumY2 = sumY2 + points2[n][1]
    c1 = int((sumX2 /