FORMAT_CONVERSION_MAXCOMPUTE/format_convert/convert_image.py

import logging
import os
import sys
sys.path.append(os.path.dirname(__file__) + "/../")
import traceback
import cv2
from format_convert import get_memory_info
from format_convert.utils import judge_error_code, get_formatted_table, add_div
from format_convert.table_correct import get_rotated_image
from format_convert.convert_need_interface import from_otr_interface, from_ocr_interface


def image_preprocess(image_np, image_path, use_ocr=True):
    logging.info("into image_preprocess")
    try:
        # 长 宽
        # resize_size = (1024, 768)
        # 限制图片大小
        # resize_image(image_path, resize_size)

        # 图片倾斜校正，写入原来的图片路径
        g_r_i = get_rotated_image(image_np, image_path)
        if g_r_i == [-1]:
            return [-1], [], [], 0

        # otr需要图片resize, 写入另一个路径
        image_np = cv2.imread(image_path)
        best_h, best_w = get_best_predict_size(image_np)
        image_resize = cv2.resize(image_np, (best_w, best_h), interpolation=cv2.INTER_AREA)
        # image_resize_path = image_path[:-4] + "_resize" + image_path[-4:]
        image_resize_path = image_path.split(".")[0] + "_resize." + image_path.split(".")[-1]
        cv2.imwrite(image_resize_path, image_resize)

        # 调用otr模型接口
        with open(image_resize_path, "rb") as f:
            image_bytes = f.read()
        points, split_lines, bboxes, outline_points = from_otr_interface(image_bytes)
        if judge_error_code(points):
            return points, [], [], 0

        # 将resize后得到的bbox根据比例还原
        ratio = (image_np.shape[0]/best_h, image_np.shape[1]/best_w)
        for i in range(len(bboxes)):
            bbox = bboxes[i]
            bboxes[i] = [(int(bbox[0][0]*ratio[1]), int(bbox[0][1]*ratio[0])),
                         (int(bbox[1][0]*ratio[1]), int(bbox[1][1]*ratio[0]))]
        for i in range(len(split_lines)):
            line = split_lines[i]
            split_lines[i] = [(int(line[0][0]*ratio[1]), int(line[0][1]*ratio[0])),
                              (int(line[1][0]*ratio[1]), int(line[1][1]*ratio[0]))]
        for i in range(len(points)):
            point = points[i]
            points[i] = (int(point[0]*ratio[1]), int(point[1]*ratio[0]))

        for i in range(len(outline_points)):
            point = outline_points[i]
            outline_points[i] = [(int(point[0][0]*ratio[1]), int(point[0][1]*ratio[0])),
                                 (int(point[1][0]*ratio[1]), int(point[1][1]*ratio[0]))]

        # 查看是否能输出正确框
        for box in bboxes:
            cv2.rectangle(image_np, box[0], box[1], (0, 255, 0), 2)
            # cv2.namedWindow('bbox', 0)
            # cv2.imshow("bbox", image_np)
            # cv2.waitKey(0)

        # 调用ocr模型接口
        with open(image_path, "rb") as f:
            image_bytes = f.read()
        # 有表格
        if len(bboxes) >= 2:
            text_list, bbox_list = from_ocr_interface(image_bytes, True)
            if judge_error_code(text_list):
                return text_list, [], [], 0

            # for i in range(len(text_list)):
            #     print(text_list[i], bbox_list[i])
            # 查看是否能输出正确框

            # for box in bbox_list:
            #     cv2.rectangle(image_np, (int(box[0][0]), int(box[0][1])),
            #                   (int(box[2][0]), int(box[2][1])), (255, 0, 0), 1)
            #     cv2.namedWindow('bbox', 0)
            #     cv2.imshow("bbox", image_np)
            #     cv2.waitKey(0)

            text, column_list = get_formatted_table(text_list, bbox_list, bboxes, split_lines)
            if judge_error_code(text):
                return text, [], [], 0
            is_table = 1
            return text, column_list, outline_points, is_table

        # 无表格
        else:
            if use_ocr:
                text = from_ocr_interface(image_bytes)
                if judge_error_code(text):
                    return text, [], [], 0

                is_table = 0
                return text, [], [], is_table
            else:
                is_table = 0
                return None, [], [], is_table

    except Exception as e:
        logging.info("image_preprocess error")
        print("image_preprocess", traceback.print_exc())
        return [-1], [], [], 0


@get_memory_info.memory_decorator
def picture2text(path, html=False):
    logging.info("into picture2text")
    try:
        # 判断图片中表格
        img = cv2.imread(path)
        if img is None:
            return [-3]

        text, column_list, outline_points, is_table = image_preprocess(img, path)
        if judge_error_code(text):
            return text

        if html:
            text = add_div(text)
        return [text]
    except Exception as e:
        logging.info("picture2text error!")
        print("picture2text", traceback.print_exc())
        return [-1]


def get_best_predict_size(image_np, times=64):
    sizes = []
    for i in range(1, 100):
        if i*times <= 3000:
            sizes.append(i*times)
    sizes.sort(key=lambda x: x, reverse=True)

    min_len = 10000
    best_height = sizes[0]
    for height in sizes:
        if abs(image_np.shape[0] - height) < min_len:
            min_len = abs(image_np.shape[0] - height)
            best_height = height

    min_len = 10000
    best_width = sizes[0]
    for width in sizes:
        if abs(image_np.shape[1] - width) < min_len:
            min_len = abs(image_np.shape[1] - width)
            best_width = width

    return best_height, best_width