predict_system.py 5.0 KB
Newer Older
L
LDOUBLEV 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import utility
from ppocr.utils.utility import initial_logger
logger = initial_logger()
import cv2
import predict_det
import predict_rec
import copy
import numpy as np
import math
import time
L
LDOUBLEV 已提交
25 26 27 28
from ppocr.utils.utility import get_image_file_list
from PIL import Image
from tools.infer.utility import draw_ocr
import os
L
LDOUBLEV 已提交
29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70


class TextSystem(object):
    def __init__(self, args):
        self.text_detector = predict_det.TextDetector(args)
        self.text_recognizer = predict_rec.TextRecognizer(args)

    def get_rotate_crop_image(self, img, points):
        img_height, img_width = img.shape[0:2]
        left = int(np.min(points[:, 0]))
        right = int(np.max(points[:, 0]))
        top = int(np.min(points[:, 1]))
        bottom = int(np.max(points[:, 1]))
        img_crop = img[top:bottom, left:right, :].copy()
        points[:, 0] = points[:, 0] - left
        points[:, 1] = points[:, 1] - top
        img_crop_width = int(np.linalg.norm(points[0] - points[1]))
        img_crop_height = int(np.linalg.norm(points[0] - points[3]))
        pts_std = np.float32([[0, 0], [img_crop_width, 0],\
            [img_crop_width, img_crop_height], [0, img_crop_height]])
        M = cv2.getPerspectiveTransform(points, pts_std)
        dst_img = cv2.warpPerspective(
            img_crop,
            M, (img_crop_width, img_crop_height),
            borderMode=cv2.BORDER_REPLICATE)
        dst_img_height, dst_img_width = dst_img.shape[0:2]
        if dst_img_height * 1.0 / dst_img_width >= 1.5:
            dst_img = np.rot90(dst_img)
        return dst_img

    def print_draw_crop_rec_res(self, img_crop_list, rec_res):
        bbox_num = len(img_crop_list)
        for bno in range(bbox_num):
            cv2.imwrite("./output/img_crop_%d.jpg" % bno, img_crop_list[bno])
            print(bno, rec_res[bno])

    def __call__(self, img):
        ori_im = img.copy()
        dt_boxes, elapse = self.text_detector(img)
        if dt_boxes is None:
            return None, None
        img_crop_list = []
71 72 73

        dt_boxes = sorted_boxes(dt_boxes)

L
LDOUBLEV 已提交
74 75 76 77 78
        for bno in range(len(dt_boxes)):
            tmp_box = copy.deepcopy(dt_boxes[bno])
            img_crop = self.get_rotate_crop_image(ori_im, tmp_box)
            img_crop_list.append(img_crop)
        rec_res, elapse = self.text_recognizer(img_crop_list)
79
        # self.print_draw_crop_rec_res(img_crop_list, rec_res)
L
LDOUBLEV 已提交
80 81 82
        return dt_boxes, rec_res


83 84 85 86
def sorted_boxes(dt_boxes):
    """
    Sort text boxes in order from top to bottom, left to right
    args:
T
tink2123 已提交
87
        dt_boxes(array):detected text boxes with shape [4, 2]
88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103
    return:
        sorted boxes(array) with shape [4, 2]
    """
    num_boxes = dt_boxes.shape[0]
    sorted_boxes = sorted(dt_boxes, key=lambda x: x[0][1])
    _boxes = list(sorted_boxes)

    for i in range(num_boxes - 1):
        if abs(_boxes[i+1][0][1] - _boxes[i][0][1]) < 10 and \
            (_boxes[i + 1][0][0] < _boxes[i][0][0]):
            tmp = _boxes[i]
            _boxes[i] = _boxes[i + 1]
            _boxes[i + 1] = tmp
    return _boxes


L
LDOUBLEV 已提交
104 105
if __name__ == "__main__":
    args = utility.parse_args()
L
LDOUBLEV 已提交
106
    image_file_list = get_image_file_list(args.image_dir)
L
LDOUBLEV 已提交
107
    text_sys = TextSystem(args)
L
LDOUBLEV 已提交
108
    is_visualize = True
L
LDOUBLEV 已提交
109 110 111 112 113 114 115 116 117 118 119 120 121
    for image_file in image_file_list:
        img = cv2.imread(image_file)
        if img is None:
            logger.info("error in loading image:{}".format(image_file))
            continue
        starttime = time.time()
        dt_boxes, rec_res = text_sys(img)
        elapse = time.time() - starttime
        print("Predict time of %s: %.3fs" % (image_file, elapse))
        dt_num = len(dt_boxes)
        dt_boxes_final = []
        for dno in range(dt_num):
            text, score = rec_res[dno]
L
LDOUBLEV 已提交
122
            if score >= 0.5:
L
LDOUBLEV 已提交
123 124 125
                text_str = "%s, %.3f" % (text, score)
                print(text_str)
                dt_boxes_final.append(dt_boxes[dno])
L
LDOUBLEV 已提交
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140

        if is_visualize:
            image = Image.fromarray(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
            boxes = dt_boxes
            txts = [rec_res[i][0] for i in range(len(rec_res))]
            scores = [rec_res[i][1] for i in range(len(rec_res))]

            draw_img = draw_ocr(
                image, boxes, txts, scores, draw_txt=True, drop_score=0.5)
            draw_img_save = "./doc/imgs_results/"
            if not os.path.exists(draw_img_save):
                os.makedirs(draw_img_save)
            cv2.imwrite(
                os.path.join(draw_img_save, os.path.basename(image_file)),
                draw_img)