---
layout: default-layout
title: TextDetectionMode - Dynamsoft Capture Vision Parameters
description: Reference for the TextDetectionMode parameter in Dynamsoft Capture Vision, which controls how text zones are detected in an image - with results used either to erase text areas before barcode/document processing (via IfEraseTextZone) or as mandatory input for text recognition tasks.
keywords: TextDetectionMode, parameter reference, parameter
---


# TextDetectionMode

Parameter `TextDetectionMode` determines how to detect the text area. For tasks like barcode reading or border detection, the text part is not important; while for the task of text recognition, the results of detected text zone are mandatory.

## JSON Structure

**Location in template:**
```
ImageParameterOptions[i]
    └── ApplicableStages[j] (Stage object where Stage = "SST_DETECT_TEXT_ZONES")
        └── TextDetectionMode
```

**Parent object:** [DetectTextZonesStage](stage-detect-text-zones.md)

**Example:**

```json
{
    "TextDetectionMode": {
        "Mode": "TTDM_LINE"
    }
}
```

> [!NOTE]
> - This snippet shows only the `TextDetectionMode` parameter.
> - To use it, embed this parameter within a [DetectTextZonesStage](stage-detect-text-zones.md) object.
> - For the complete JSON structure, see:
>   - [Full JSON Structure]({{ site.dcvb_parameters }}file/index.html#full-json-structure)
>   - [Minimal Valid JSON]({{ site.dcvb_parameters }}file/index.html#minimal-valid-json-example)

## Parameter Details

Parameter `TextDetectionMode` consist of a group of text detection mode objects. Each text detection mode object includes a candidate mode and a series of mode arguments. The mode arguments of the text detection mode object is shown as follow:

### Mode Arguments

<table style = "text-align:left">
    <thead>
        <tr>
            <th nowrap="nowrap">Mode Argument Name</th>
            <th nowrap="nowrap">Mode Argument Summary</th>
        </tr>
    </thead>
    <tr>
        <td rowspan = "3" style="vertical-align:text-top">Mode<br>(Required)</td>
        <td><b>Description</b><br>Any one in Candidate Mode List as string
        </td>
    </tr>
    <tr>
        <td><b>Type</b><br><i>String</i>
        </td>
    </tr>
    <tr>
        <td><b>Candidate Mode List</b><br>TTDM_WORD
            <br>TTDM_LINE
            <br>TTDM_LAYOUT (Not supported yet)
            <br>TTDM_SKIP
        </td>
    </tr>
    <tr>
        <td rowspan = "5" style="vertical-align:text-top">Direction<br>(Optional)</td>
        <td><b>Description</b><br>Set the direction to perform text zone detection.
        </td>
    </tr>
    <tr>
        <td><b>Type</b><br><i>String</i>
        </td>
    </tr>
    <tr>
        <td><b>Range</b><br>"HORIZONTAL", "VERTICAL", "OBLIQUE", "UNKNOWN"
        </td>
    </tr>
    <tr>
        <td><b>Default Value</b><br>"UNKNOWN"
        </td>
    </tr>
    <tr>
        <td><b>Valid For</b><br>All modes
        </td>
    </tr>
    <tr>
        <td rowspan = "4" style="vertical-align:text-top">CharHeightRange<br>(Optional)</td>
        <td><b>Description</b><br>Sets the range of letter height in pixel or a percentage.
        <br>Format: [MinHeight, MaxHeight, ByThousandth].
        <br>if ByThousandth=1, the allowed values for MinHeight/MaxHeight=[1, 1000]
        <br>if ByThousandth=0, the allowed values for MinHeight/MaxHeight=[1, 0x7fffffff]
        </td>
    </tr>
    <tr>
        <td><b>Type</b><br><i>int array</i>
        </td>
    </tr>
    <tr>
        <td><b>Range</b><br>[1, 0x7fffffff]
        </td>
    </tr>
    <tr>
        <td><b>Valid For</b><br>All modes
        </td>
    </tr>
    <tr>
        <td rowspan = "6" style="vertical-align:text-top">MaxSpacingInALine<br>(Optional)</td>
        <td><b>Description</b><br>Sets the maximum spacing between characters treated as one line.
        </td>
    </tr>
    <tr>
        <td><b>Type</b><br><i>int</i>
        </td>
    </tr>
    <tr>
        <td><b>Range</b><br>[-1, 0x7fffffff]
        </td>
    </tr>
    <tr>
        <td><b>Default Value</b><br>-1
        </td>
    </tr>
    <tr>
        <td><b>Valid For</b><br>TTDM_LINE
        </td>
    </tr>
    <tr>
        <td><b>Remarks</b><br>-1: means automatically set by the library.<br>It is a percentage value relative to the average letter height of each line.
        </td>
    </tr>
    <tr>
        <td rowspan = "4" style="vertical-align:text-top">Sensitivity<br>(Optional)</td>
        <td><b>Description</b><br>Sets the sensitivity of text detection.
        </td>
    </tr>
    <tr>
        <td><b>Type</b><br><i>int</i>
        </td>
    </tr>
    <tr>
        <td><b>Range</b><br>[1, 9]
        </td>
    </tr>
    <tr>
        <td><b>Valid For</b><br>All modes
        </td>
    </tr>
        <tr>
        <td rowspan = "5" style="vertical-align:text-top">StringLengthRange<br>(Optional)</td>
        <td><b>Description</b><br>Sets a range to restrict the string length.
        </td>
    </tr>
    <tr>
        <td><b>Type</b><br><i>int[]</i>
        </td>
    </tr>
    <tr>
        <td><b>Range</b><br>Each member of the array should be in [0, 0x7fffffff]
        </td>
    </tr>
    <tr>
        <td><b>Default Value</b><br>LabelRecognizer: [3,50]. <br>BarcodeReader & DocumentNormalizer: null
        </td>
    </tr>
    <tr>
        <td><b>Valid For</b><br>All modes
        </td>
    </tr>
    <tr>
        <td rowspan = "5" style="vertical-align:text-top">LibraryFileName<br>(Optional)</td>
        <td><b>Description</b><br>Sets the file name of the library to load dynamically.
        </td>
    </tr>
    <tr>
        <td><b>Type</b><br><i>String</i>
        </td>
    </tr>
    <tr>
        <td><b>Range</b><br>A string value representing file name.
        </td>
    </tr>
    <tr>
        <td><b>Default Value</b><br>""
        </td>
    </tr>
    <tr>
        <td><b>Valid For</b><br>All modes.
        </td>
    </tr>
    <tr>
        <td rowspan = "5" style="vertical-align:text-top">LibraryParameters<br>(Optional)</td>
        <td><b>Description</b><br>The library must be in the same place with Dynamsoft Barcode Reader Library.
        </td>
    </tr>
    <tr>
        <td><b>Type</b><br><i>String</i>
        </td>
    </tr>
    <tr>
        <td><b>Range</b><br>A string value representing parameters.
        </td>
    </tr>
    <tr>
        <td><b>Default Value</b><br>""
        </td>
    </tr>
    <tr>
        <td><b>Valid For</b><br>All modes.
        </td>
    </tr>
</table>

### Default Setting

#### For Barcode Decoding

```json
{
    "TextDetectionMode" : 
    {
        "Mode" : "TTDM_LINE",
        "Direction" : "UNKNOWN",
        "CharHeightRange" : [1, 1000, 1],
        "MaxSpacingInALine" : -1,
        "Sensitivity" : 3
    }
}
```

#### For Label Recognizing

```json
{
    "TextDetectionMode" : 
    {
        "Mode" : "TTDM_LINE",
        "Direction" : "HORIZONTAL",
        "CharHeightRange" : [20, 1000, 1],
        "MaxSpacingInALine" : -1,
        "Sensitivity" : 7
    }
}
```

#### For Document Scanning

```json
{
    "TextDetectionMode" : 
    {
        "Mode" : "TTDM_WORD",
        "Direction" : "HORIZONTAL",
        "CharHeightRange" : [1, 1000, 1],
        "Sensitivity" : 7
    }
}
```

## Candidate Mode Introductions

### TTDM_WORD

Find the text area based on the "word" objects. It is the fastest text detection mode but has the lowest accuracy.

### TTDM_LINE

Find the text area based on the "text line" objects. This mode has higher accuracy than the `TTDM_WORD`. It is currently implemented as the default mode for text line recognizing.

### TTDM_LAYOUT

Find the text area based on the layout. A mode with even higher accuracy than the `TTDM_LINE`. Not supported yet.