表格识别算法ch_ppstructure_openatom_SLANetv2_infer报错

[Wangqi12138]

🐛 Bug (问题描述)

使用PPStructure表格识别ch_ppstructure_openatom_SLANetv2_infer算法，输出都是空白，是什么原因？PPOCRv4版本

🏃‍♂️ Environment (运行环境)

python 3.10
paddlepaddle 2.6.1
paddleocr 2.7.2

🌰 Minimal Reproducible Example (最小可复现问题的Demo)

models_dir = "/data/models/paddle_ocr/"
rec_model_dir = models_dir + "ch_PP-OCRv4_rec_infer"
det_model_dir = models_dir + "ch_PP-OCRv4_det_infer"
table_model_dir = models_dir + "ch_ppstructure_openatom_SLANetv2_infer"
layout_model_dir = models_dir + "picodet_lcnet_x1_0_fgd_layout_table_infer"

table_engine = PPStructure(show_log=False,table_model_dir = table_model_dir,layout_model_dir = layout_model_dir,det_model_dir = det_model_dir,
rec_model_dir = rec_model_dir)
image = "/data/uploaded_images/cropped_image_2200.614990234375.jpg"
img = cv2.imread(image)
result = table_engine(img)
print(result)

[GreatV]

试试paddleocr 2.8版本

[ViCtOr-dev13]

🐛 Bug (问题描述)

🏃‍♂️Environment (运行环境)

python 3.10
paddleocr 2.9.1
paddlepaddle 0.0.0 (for Mac but work fine)

🌰 Reproducible Example (最小可复现问题的Demo)

Import

from paddleocr import PPStructure,draw_structure_result,save_structure_res
import cv2
import os
import numpy as np
from PIL import Image

PPstructure

table_engine = PPStructure(type = "structure", lang="en",det_model_dir="en_ppocr_mobile_v2.0_table_det_infer",rec_model_dir="en_ppocr_mobile_v2.0_table_rec_infer") # Initialize the table structure recognition engine

show_log

table_model_dir='..../table/en_ppstructure_mobile_v2.0_SLANet_infer
table_char_dict_path='.../lib/python3.10/site-packages/paddleocr/ppocr/utils/dict/table_structure_dict.txt'
layout_model_dir='..../layout/picodet_lcnet_x1_0_fgd_layout_infer'

Code

save_folder = './output'
img_path = 'capture.png'
image = cv2.imread(img_path)
image_array = np.array(image)
result = table_engine(image)
save_structure_res(result, save_folder,os.path.basename(img_path).split('.')[0])
for line in result:
    line.pop('img')
    print(line)



font_path = 'PaddleOCR/doc/fonts/simfang.ttf' # font provieded in PaddleOCR
image = Image.open(img_path).convert('RGB')
im_show = draw_structure_result(image, result,font_path=font_path)
im_show = Image.fromarray(im_show)
im_show.save('result.jpg')

📷 capture.png

[GreatV]

@ViCtOr-dev13, please open a new issue.