Description
Table structure recognition based on hocr with Tesseract architecture, for PDF documents.
Tesseract has been trained on a variety of datasets to improve its recognition capabilities. These datasets include images of text in various languages and scripts, as well as images with different font styles, sizes, and orientations. The training process involves feeding the engine with a large number of images and their corresponding text, allowing the engine to learn the patterns and characteristics of different text styles. One of the most important datasets used in training Tesseract is the UNLV dataset, which contains over 400,000 images of text in different languages, scripts, and font styles. This dataset is widely used in the OCR community and has been instrumental in improving the accuracy of Tesseract. Other datasets that have been used in training Tesseract include the ICDAR dataset, the IIIT-HWS dataset, and the RRC-GV-WS dataset.
In addition to these datasets, Tesseract also uses a technique called adaptive training, where the engine can continuously improve its recognition capabilities by learning from new images and text. This allows Tesseract to adapt to new text styles and languages, and improve its overall accuracy.
Predicted Entities
How to use
pdf_to_hocr = PdfToHocr() \
.setInputCol("content") \
.setOutputCol("hocr")
tokenizer = HocrTokenizer() \
.setInputCol("hocr") \
.setOutputCol("token") \
pdf_to_image = PdfToImage() \
.setInputCol("content") \
.setOutputCol("image") \
.setPageNumCol("tmp_pagenum") \
.setImageType(ImageType.TYPE_3BYTE_BGR)
table_detector = ImageTableDetector \
.pretrained("general_model_table_detection_v2", "en", "public/ocr/models") \
.setInputCol("image") \
.setOutputCol("table_regions") \
.setScoreThreshold(0.9) \
.setApplyCorrection(True) \
.setScaleWidthToCol("width_dimension") \
.setScaleHeightToCol("height_dimension")
image_scaler = ImageScaler() \
.setWidthCol("width_dimension") \
.setHeightCol("height_dimension")
hocr_to_table = HocrToTextTable() \
.setInputCol("hocr") \
.setRegionCol("table_regions") \
.setOutputCol("tables")
draw_annotations = ImageDrawAnnotations() \
.setInputCol("scaled_image") \
.setInputChunksCol("tables") \
.setOutputCol("image_with_annotations") \
.setFilledRect(False) \
.setFontSize(5) \
.setRectColor(Color.red)
draw_regions = ImageDrawRegions() \
.setInputCol("scaled_image") \
.setInputRegionsCol("table_regions") \
.setOutputCol("image_with_regions") \
.setRectColor(Color.red)
pipeline1 = PipelineModel(stages=[
pdf_to_hocr,
tokenizer,
pdf_to_image,
table_detector,
image_scaler,
draw_regions,
hocr_to_table
])
test_image_path = "data/pdfs/f1120.pdf"
bin_df = spark.read.format("binaryFile").load(test_image_path)
result = pipeline1.transform(bin_df).cache().drop("tmp_pagenum")
result = result.filter(result.pagenum == 1)
val pdf_to_hocr = new PdfToHocr()
.setInputCol("content")
.setOutputCol("hocr")
val tokenizer = new HocrTokenizer()
.setInputCol("hocr")
.setOutputCol("token")
val pdf_to_image = new PdfToImage()
.setInputCol("content")
.setOutputCol("image")
.setPageNumCol("tmp_pagenum")
.setImageType(ImageType.TYPE_3BYTE_BGR)
val table_detector = ImageTableDetector
.pretrained("general_model_table_detection_v2", "en", "public/ocr/models")
.setInputCol("image")
.setOutputCol("table_regions")
.setScoreThreshold(0.9)
.setApplyCorrection(True)
.setScaleWidthToCol("width_dimension")
.setScaleHeightToCol("height_dimension")
val image_scaler = new ImageScaler()
.setWidthCol("width_dimension")
.setHeightCol("height_dimension")
val hocr_to_table = new HocrToTextTable()
.setInputCol("hocr")
.setRegionCol("table_regions")
.setOutputCol("tables")
val draw_annotations = new ImageDrawAnnotations()
.setInputCol("scaled_image")
.setInputChunksCol("tables")
.setOutputCol("image_with_annotations")
.setFilledRect(False)
.setFontSize(5)
.setRectColor(Color.red)
val draw_regions = new ImageDrawRegions()
.setInputCol("scaled_image")
.setInputRegionsCol("table_regions")
.setOutputCol("image_with_regions")
.setRectColor(Color.red)
val pipeline1 = new PipelineModel().setStages(Array(
pdf_to_hocr,
tokenizer,
pdf_to_image,
table_detector,
image_scaler,
draw_regions,
hocr_to_table))
val test_image_path = "data/pdfs/f1120.pdf"
val bin_df = spark.read.format("binaryFile").load(test_image_path)
val result = pipeline1.transform(bin_df).cache().drop("tmp_pagenum")
result = result.filter(result.pagenum == 1)
Example
Output text
path modificationTime length hocr height_dimension width_dimension pagenum token image total_pages documentnum table_regions scaled_image image_with_regions tables exception table_index
file:/content/f11... 2023-01-23 08:16:... 3471478 <div title="bbox ... 791 611 1 [{token, 0, 2, 23... {file:/content/f1... 1 0 {0, 0, 32.839153,... {file:/content/f1... {file:/content/f1... {0, 0, 0.0, 0.0,... null 0
Filename: f1120.pdf
Page: 1
Table: 0
5
col0 col1 col2 col3 col4
0 Schedule C Dividends instructions , ) Inclusions , and Sp... ( a ) Dividends inclusions and ( b ) % ( c ) Special ( a ) × deductions ( b )
1 1 Dividends from less - than - 20 % - owned dome... 234 50 None
2 2 Dividends from 20 % - or - more - owned domest... 324123 65 None
3 3 Dividends on certain debt - financed stock of ... 324 instructions see None
4 4 Dividends on certain preferred stock of less -... 234 23 . 3 None
5 5 Dividends on certain preferred stock of 20 % -... 42134 26 . 7 None
6 6 Dividends from less - than - 20 % - owned fore... 4234 50 None
7 7 Dividends from 20 % - or - more - owned foreig... 4234 65 None
8 8 Dividends from wholly owned foreign subsidiaries 42348987 100 None
9 9 Subtotal . Add lines 1 through 8 . See instruc... 987 instructions see None
10 10 Dividends from domestic corporations received ... 9786 100 None
11 11 Dividends from affiliated group members . 789 100 None
12 12 Dividends from certain FSCs 0.00 100 None
13 13 Foreign - source portion of dividends received... 421.34 100 None
14 14 Dividends from foreign corporations not includ... 2341.23 None None
15 15 Section 965 ( a ) inclusion . 1234.14 instructions see None
16 16a Subpart F inclusions derived from the sale by ... 46.54 100 None
17 b Subpart F inclusions derived from hybrid divid... 6453.65 None None
18 c Other ( attach inclusions Form ( s ) 5471 from... 985.76 None None
19 17 Global Intangible Low - Taxed Income ( GILTI )... 23.41 None None
20 18 Gross - up for foreign taxes deemed paid 1.01 None None
21 19 IC - DISC and former DISC dividends not includ... 3123.91 None None
22 20 Other dividends 12.23 None None
23 21 Deduction for dividends paid on certain prefer... 1.23
24 22 Section 250 deduction ( attach Form 8993 ) 3.41
25 23 Total dividends and inclusions . Add column ( ... 2341.23 None None
Model Information
| Model Name: | hocr_table_recognition_pdf |
| Type: | ocr |
| Compatibility: | Visual NLP 4.2.4+ |
| License: | Licensed |
| Edition: | Official |
| Language: | en |