nsq_convert2txt/file_processing/models/picture_extract.py

# coding:utf-8
from PIL import Image
import os


def extract_image(xObject, base_dir, image_name):
    try:
        if xObject.attrs["Subtype"].name == 'Image':
            size = (xObject.attrs["Width"], xObject.attrs["Height"])
            data = xObject.get_data()
            if xObject.attrs["ColorSpace"].name == 'DeviceRGB':
                mode = "RGB"
            else:
                mode = "P"
            # 保存图片的文件名前缀
            img_pre = os.path.join(base_dir, image_name)
            if 'Filter' in xObject.attrs:
                if xObject.attrs["Filter"].name == 'FlateDecode':
                    img = Image.frombytes(mode, size, data)
                    img.save(img_pre + ".png")
                    return img_pre + ".png"
                elif xObject.attrs["Filter"].name == 'DCTDecode':
                    img = open(img_pre + ".jpg", "wb")
                    img.write(data)
                    img.close()
                    return img_pre + ".jpg"
                elif xObject.attrs["Filter"].name == 'JPXDecode':
                    img = open(img_pre + ".jp2", "wb")
                    img.write(data)
                    img.close()
                    return img_pre + ".jp2"
                elif xObject.attrs["Filter"].name == 'CCITTFaxDecode':
                    img = open(img_pre + ".tiff", "wb")
                    img.write(data)
                    img.close()
                    return img_pre + ".tiff"
            else:
                img = Image.frombytes(mode, size, data)
                img.save(img_pre + ".png")
                return img_pre + ".png"
    except Exception as e:
        # print(f"图片提取失败-->{e}")
        return None


class Picture(object):
    def __init__(self, min_y, max_y, height, width, image_path):
        self.min_y = min_y
        self.max_y = max_y
        self.width = width
        self.height = height
        self.image_path = image_path
        self.content = ""