chenguilong
/
ocr-table


			
							123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180
							# -*- coding: UTF-8 -*-
import json
from base64 import b64decode

import cv2
import numpy as np
from fastapi import FastAPI, Request
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
from paddleocr import PaddleOCR, PPStructure
from sx_utils.sxweb import *
from sx_utils.sximage import *
import threading
import os
import re
from sx_utils.sx_log import *
import paddleclas

from cores.post_hander import *


format_print()

# 初始化APP
app = FastAPI()
origins = ["*"]

app.add_middleware(
    CORSMiddleware,
    allow_origins=origins,
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)


table_engine_lock = threading.Lock()

table_engine = PPStructure(layout=False,
                           table=True,
                           use_gpu=True,
                           show_log=True,
                           use_angle_cls=True,
                        #    det_model_dir="models/det/det_table_v2",
                        #    det_model_dir="models/det/det_table_v3",
                        #    rec_model_dir="models/rec/rec_table_v1",
                           table_model_dir="models/table/SLANet_911")


cls_lock = threading.Lock()

cls_model = paddleclas.PaddleClas(model_name="text_image_orientation")
# # 普通表格
# table_engine = PPStructure(layout=False,
#                            table=True,
#                            use_gpu=use_gpu,
#                            show_log=True,
#                            det_model_dir="models/det/det_table_v2",
#                            rec_model_dir="./models/rec/rec_table_v1",
#                            table_model_dir="models/table/SLANet_v2")
#
# # 长度较长表格
# table_engine1 = PPStructure(layout=False,
#                             table=True,
#                             use_gpu=use_gpu,
#                             show_log=True,
#                             det_model_dir="models/det/det_table_v1",
#                             rec_model_dir="./models/rec/rec_table_v1",
#                             table_model_dir="./models/table/SLAnet_v1")
#
# # 针对某些特殊情况的补充模型
# table_engine2 = PPStructure(layout=False,
#                             table=True,
#                             use_gpu=use_gpu,
#                             show_log=True,
#                             det_model_dir="models/det/det_table_v3",
#                             rec_model_dir="./models/rec/rec_table_v1",
#                             table_model_dir="./models/table/SLAnet_v1")
#
#
#

# 用于判断各个角度table的识别效果，识别的字段越多，效果越好
def cal_html_to_chs(html):
    res = []
    rows = re.split('<tr>', html)
    for row in rows:
        row = re.split('<td>', row)
        cells = list(map(lambda x: x.replace('</td>', '').replace('</tr>', ''), row))
        rec_str = ''.join(cells)
        for tag in ['<html>', '</html>', '<body>', '</body>', '<table>', '</table>', '<tbody>', '</tbody>']:
            rec_str = rec_str.replace(tag, '')

        res.append(rec_str)

    rec_res = ''.join(res).replace(' ', '')
    rec_res = re.split('<tdcolspan="\w+">', rec_res)
    rec_res = ''.join(rec_res).replace(' ', '')
    print(rec_res)
    return len(rec_res)


def predict_cls(image, conf=0.8):
    try:
        cls_lock.acquire()
        result = cls_model.predict(image)
    finally:
        cls_lock.release()
    for res in result:
        score = res[0]['scores'][0]
        label_name = res[0]['label_names'][0]
        print(f"score: {score}, label_name: {label_name}")
        if score > conf:
            return int(label_name)
    return -1


def rotate_to_zero(image, current_degree):
    current_degree = current_degree // 90
    if current_degree == 0:
        return image
    to_rotate = (4 - current_degree) - 1
    image = cv2.rotate(image, to_rotate)
    return image


def get_zero_degree_image(img):
    step = 0.2
    for idx, i in enumerate([-1, 0, 1, 2]):
        if i >= 0:
            image = cv2.rotate(img.copy(), i)
        else:
            image = img.copy()
        conf = 0.8 - (idx * step)
        current_degree = predict_cls(image, conf)  # 0 90  180 270 -1 识别不出来
        if current_degree != -1:
            img = rotate_to_zero(image, current_degree)
            break
        else:
            continue
    return img

def table_res(im, ROTATE=-1):
    im = im.copy()
    # 获取正向图片
    img = get_zero_degree_image(im)
    # cv2.imwrite('1.jpg', img)
    try:
        table_engine_lock.acquire()
        res = table_engine(img)
    finally:
        table_engine_lock.release()
    html = res[0]['res']['html']
    return res, html

class TableInfo(BaseModel):
    image: str
    det: str


@app.get("/ping")
def ping():
    return 'pong!!!!!!!!!'


@app.post("/ocr_system/table")
@web_try()
def table(image: TableInfo):
    img = base64_to_np(image.image)
    res, html = table_res(img)

    if html:
        post_hander = PostHandler(html)
        return {'html': post_hander.format_predict_html}
    else:
        raise Exception('无法识别')


print('table system init success!')