See all results >
Table of contents

The following diagram shows how sections connect to each other to form tasks:

1.Region Pre-Detection
2.1.Shared Detection
2.2.Barcode Localization
2.2.Text-line Localization
2.2.Document Detection
3.Barcode Decoding
3.Text-line Recognition
3.Document Normalization

In this article, we’ll discuss the section Text-line Recognition which is usually the 3rd section of a “Recognize-Text-Lines” task.

Section 3 - Text-line Recognition

The purpose of this section is to recognize the text from the text-line areas identified in the previous section “Text-line Localization”.

Constituting Stages

This section consists of the following stages:

  1. Cropping: to cut out the text-line areas based on text-line localization results. This results in one or multiple colour images.
  2. Grayscaling: to convert the colour image(s) to grayscale.
  3. Transforming: to transform the grayscale image(s).
  4. Text-line-recognizing: to recognize the text.

Output and Parameters

Each of these stages has its own output (known as an intermediate result) and usually a specific parameter that can regulate the operation:

Stage Intermediate Result Type Related Parameter
Cropping IRUT_COLOUR_IMAGE N/A
Grayscaling IRUT_GRAYSCALE_IMAGE ColourConversionModes
Transforming IRUT_TRANSFORMED_GRAYSCALE_IMAGE GrayscaleTransformationModes
Text-line-recognizing IRUT_RECOGNIZED_TEXT_LINES DictionaryPath
DictionaryCorrectionThresholds
StringLengthRange
StringRegExPattern

This page is compatible for:

Is this page helpful?

YesYes NoNo

In this article: