numpy数组做图片拼接的实现（concatenate、vstack、hstack）

2019-11-25 11:27:36

字体：大中小

来源：转载

供稿：网友

两种方法拼接

#img = np.vstack((img, img2))  # vstack按垂直方向，hstack按水平方向img = np.concatenate((img, img2), axis=0)  # axis=0 按垂直方向，axis=1 按水平方向

统一图片大小，保证数组维度一致避免拼接失败。把图片全部调整成第一张图的宽高

def img_size(image_names,width, height):  for i in image_names:    img = cv2.imread(os.path.join(img_path, i))    img_resize = cv2.resize(img, (width, height), interpolation=cv2.INTER_CUBIC)    cv2.imwrite(os.path.join(img_path, i), img_resize)    print(os.path.join(img_path, i))

完整案例，拼接文件夹中的所有图片

import cv2import osimport numpy as npdef img_size(image_names,width, height):  for i in image_names:    img = cv2.imread(os.path.join(img_path, i))    img_resize = cv2.resize(img, (width, height), interpolation=cv2.INTER_CUBIC)    cv2.imwrite(os.path.join(img_path, i), img_resize)    print(os.path.join(img_path, i))if __name__ == '__main__':  img_path = r'F:/studytest'  image_names = [name for name in os.listdir(img_path) if os.path.splitext(name)[1] == ".jpg"]  img1 = cv2.imread(os.path.join(img_path, image_names[0]))  width, height = img1.shape[:2][::-1]  img_size(image_names,width, height)  img = img1  for i in range(1,len(image_names)):    img_page = image_names[i]    img2 = cv2.imread(os.path.join(img_path, img_page))    #img = np.vstack((img, img2))  # vstack按垂直方向，hstack按水平方向    img = np.concatenate((img, img2), axis=0)  # axis=0 按垂直方向，axis=1 按水平方向  cv2.imwrite(os.path.join(img_path,"res.jpg"), img)  # cv2.imshow("img",img)  # cv2.waitKey()``

以上就是本文的全部内容，希望对大家的学习有所帮助，也希望大家多多支持武林网。

上一篇：python 基于dlib库的人脸检测的实现

下一篇：python实现身份证实名认证的方法实例