将图片排版至docx文档中

2019年11月14号

作者：杨轮飞

前几天遇到一个小问题，收到了大概一百五十张大小不一致的照片，要将其放入到docx文档中，如果一张一张放进去，还得一张一张的去调图片大小，实在是太麻烦了，写个处理排版的代码就一劳永逸了~

先导入库：进度条、opencv、os和numpy

from tqdm import trange
import cv2
import os
import numpy as np

思路是将四张图片排在一张A4大小的背景图里，所以先写一个能够返回四个图像在背景上的左上角坐标，输出参数分别是背景图像和目标图像的行数和列数

def image_site(back, up):
    height, weight = back
    row, col = up
    pointy1 = int((height/2 - row)/2)
    pointx1 = int((weight/2 - col)/2)

    return [(pointx1, pointy1), (col + pointx1 * 3, pointy1), (pointx1, row + pointy1*3), (col + pointx1 * 3, row + pointy1*3)]

在接下来写一个将图片放入背景图的函数

def image_add(background, images, points, imagesize):
    row, col = imagesize
    for number in range(len(images)):
        point = points[number]
        background[point[1]:point[1]+row, point[0]:point[0]+col] = images[number]
    return background

每处理好四张图像就需要放入背景图中，所以再写一个创建空白背景图的函数

# BGR
def create_image(row, col, color):
    image = np.zeros([row, col, 3], np.uint8)
    image[:, :, 0] = np.zeros([row, col]) + color[0]
    image[:, :, 1] = np.zeros([row, col]) + color[1]
    image[:, :, 2] = np.zeros([row, col]) + color[2]
    return image

接下来就是主函数了，每读取四张图像，就将其添加到背景图中，并保存，该文件执行后生成一些合成图像

if __name__ == '__main__':
    file = 'card'
    endfile = 'result'
    filelist = os.listdir(file)
    color = (255, 255, 255)
    height, weight = 835, 595
    row, col = 395, 280
    points = image_site((height, weight), (row, col))
    print("标点位置:",points)

    image_PDF = []
    for number in trange(0,len(filelist)-3):
        if (number+4) % 4 == 0:
            background = create_image(height, weight, color)
            images = []
            for image_min_number in range(4):
                img = cv2.imdecode(np.fromfile(file + '\\' + filelist[image_min_number+number], dtype=np.uint8), 1)
                img = cv2.resize(img, (col, row), interpolation=cv2.INTER_AREA)
                images.append(img)
            result = image_add(background, images, points, (row, col))
            #result = cv2.resize(result, (350, 500), interpolation=cv2.INTER_AREA)
            cv2.imwrite(endfile+'\\'+str(number)+'.jpg',result,[int(cv2.IMWRITE_JPEG_QUALITY),100])
            image_PDF.append(result)
            #cv2.imshow("111", result)
            #cv2.waitKey()
    cv2.waitKey(10)
    print("合成纸张数量:",len(image_PDF))

生成图像后就简单多了，只要将其添加到文档中就行了，因为没有找到修改docx文档的页边距的函数，所以只能退而求其次，创建一个docx文档，设置好页边距后进行添加了。

from docx import Document
from tqdm import trange
import os
#
def main():
    document = Document('demo.docx')
    file = 'result'
    filelist = os.listdir(file)
    for t in trange(len(filelist)):
        document.add_picture(file+'\\'+filelist[t])
    document.save('demo.docx')
if __name__ == '__main__':
    main()

最终效果如下