# TextLineSpecification Parameters The `TextLineSpecification` object defines configurations for specified text lines, including character models, recognition patterns, and output settings. ## Example JSON ```json { "TextLineSpecificationOptions": [ { "Name": "tls_default", "BaseTextLineSpecificationName": "", "ApplicableTextLineNumbers": [0], "CharacterModelName": "", "TextLineRecModelName": "", "CharHeightRange": [5, 1000], "CharacterNormalizationModes": [], "BinarizationModes": [], "GrayscaleEnhancementModes": [], "StringLengthRange": [3, 200], "StringRegexPattern": "", "OutputResults": 1, "ConcatResults": 0, "ConcatSeparator": "", "ConcatStringLengthRange": [0, 0], "ConcatStringRegexPattern": "", "ExpectedGroupsCount": 1, "SubGroups": [], "ReferenceGroupName": "", "TextLinesCount": 1, "Position": { "Left": -1, "Top": -1, "Right": -1, "Bottom": -1 }, "ConfusableCharactersCorrection": { "ConfusionSet": "", "FontName": "", "Height": 0, "Width": 0 } } ] } ``` ## Hierarchical Structure This tree shows one `TextLineSpecification` object inside `TextLineSpecificationOptions`. ```text TextLineSpecification ├── Name ├── BaseTextLineSpecificationName ├── ApplicableTextLineNumbers ├── CharacterModelName ├── TextLineRecModelName ├── CharHeightRange ├── CharacterNormalizationModes ├── BinarizationModes ├── GrayscaleEnhancementModes ├── StringLengthRange ├── StringRegexPattern ├── OutputResults ├── ConcatResults ├── ConcatSeparator ├── ConcatStringLengthRange ├── ConcatStringRegexPattern ├── ExpectedGroupsCount ├── SubGroups ├── ReferenceGroupName ├── TextLinesCount ├── Position └── ConfusableCharactersCorrection ``` ## Top-Level Parameters | Parameter Name | Description | |:---------------|:------------| | [`Name`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/name.md) | The unique name of the TextLineSpecification object. | | [`BaseTextLineSpecificationName`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/base-text-line-specification-name.md) | The name of another TextLineSpecification to inherit from. | | [`ApplicableTextLineNumbers`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/applicable-text-line-numbers.md) | The text line numbers this specification applies to. | | [`CharacterModelName`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/character-model-name.md) | The name of the character recognition model. | | [`TextLineRecModelName`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/text-line-rec-model-name.md) | The name of the text line recognition model. | | [`CharHeightRange`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/char-height-range.md) | The expected character height range. | | [`CharacterNormalizationModes`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/character-normalization-modes.md) | The modes for character normalization. | | [`BinarizationModes`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/binarization-modes.md) | The modes for binarization. | | [`GrayscaleEnhancementModes`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/grayscale-enhancement-modes.md) | The modes for grayscale enhancement. | | [`StringLengthRange`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/string-length-range.md) | The expected string length range. | | [`StringRegexPattern`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/string-regex-pattern.md) | The regex pattern for validating recognized strings. | | [`OutputResults`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/output-results.md) | Whether to output results for this specification. | | [`ConcatResults`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/concat-results.md) | Whether to concatenate results. | | [`ConcatSeparator`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/concat-separator.md) | The separator for concatenated results. | | [`ConcatStringLengthRange`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/concat-string-length-range.md) | The length range for concatenated strings. | | [`ConcatStringRegexPattern`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/concat-string-regex-pattern.md) | The regex pattern for concatenated strings. | | [`ExpectedGroupsCount`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/expected-groups-count.md) | The expected number of text line groups. | | [`SubGroups`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/sub-groups.md) | The sub-group definitions. | | [`ReferenceGroupName`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/reference-group-name.md) | The name of the reference group. | | [`TextLinesCount`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/text-lines-count.md) | The expected number of text lines. | | [`Position`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/position.md) | The position of the text line. | | [`ConfusableCharactersCorrection`](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/text-line-specification/confusable-characters-correction.md) | The settings for confusable characters correction. | ## Usage Instructions ### Design and Configuration Strategy `TextLineSpecification` is used to apply specialized recognition settings to selected text lines. - `ApplicableTextLineNumbers` controls which text lines are targeted. - `CharacterModelName` and `TextLineRecModelName` select the CNN models used for recognition. - Other parameters define preprocessing, validation, grouping, and output behavior for the targeted lines. When multiple `TextLineSpecification` objects exist, each object affects only the lines it targets. ### Select Models Use `CharacterModelName` and `TextLineRecModelName` to bind the corresponding models to the target text lines. For model configuration details, see [CaptureVisionModel](https://www.dynamsoft.com/capture-vision/docs/core/parameters/reference/capture-vision-model/index.md). ### Select Target Text Lines and Area `ApplicableTextLineNumbers` defines which text lines use the current `TextLineSpecification` settings. - If `ApplicableTextLineNumbers` is null, all text lines use default settings. - If only some line numbers are specified, unspecified lines still use default settings. You can further constrain target lines by area using `FirstPoint`, `SecondPoint`, `ThirdPoint`, and `FourthPoint`. The final targeting result is determined by both line-number selection and area selection. For example:

text-line-specification

Example Text Line Specification
You can use the following parameters to process the above image: ```json { "ApplicableTextLineNumbers": [7, 8], "FirstPoint": [0, 60], "SecondPoint": [70, 60], "ThirdPoint": [70, 100], "FourthPoint": [0, 100] } ``` If `ApplicableTextLineNumbers` is set to cover lines 1 to 8, but only lines 7 and 8 are inside the configured area, lines 1 to 6 are excluded by area filtering. ### Configure Image Processing Modes `GrayscaleEnhancementModes` improve grayscale quality before recognition. `BinarizationModes` determine binary-image quality and directly affect character visibility in text areas before recognition. `CharacterNormalizationModes` are additional improvements, commonly based on morphological transformations. You can learn more from Image-Processing in OpenCV - Morphological Transformations. ### Quick Settings Based on an existing `TextLineSpecification` object, you can set `BaseTextLineSpecificationName` and override only the required fields. For example: ```json { "TextLineSpecificationOptions": [ { "Name": "LS_0", "CharacterModelName": "NumberLetterCharRecognition", "ApplicableTextLineNumbers": [1, 2, 3], "BinarizationModes": [ { "Mode": "BM_LOCAL_BLOCK", "BlockSizeX": 5, "BlockSizeY": 5 } ], "CharacterNormalizationModes": [ { "Mode": "CNM_MORPH", "MorphOperation": "Close", "MorphArgument": "3" } ], "StringLengthRange": [44, 44], "StringRegexPattern": "Dynamsoft", "CharHeightRange": [800, 1000] }, { "Name": "LS_1", "BaseTextLineSpecificationName": "LS_0", "ApplicableTextLineNumbers": [4, 5, 6, 7], "CharHeightRange": [600, 800] } ] } ``` In this example, `LS_1` reuses the core settings from `LS_0` and overrides only line targeting and character-height range. ### Result Composition and Output #### Result Concatenation When multiple text lines need to be combined into a single result, use the following parameters: - **`ConcatResults`**: Set to 1 to enable concatenation of multiple text line results into a single output. - **`ConcatSeparator`**: Defines the separator string used between concatenated text lines (e.g., "\n" for newline). - **`ConcatStringLengthRange`**: Specifies the valid length range [min, max] for the concatenated result. - **`ConcatStringRegexPattern`**: Regular expression pattern that the concatenated result must match. #### Grouping and Filtering - **`ExpectedGroupsCount`**: Specifies the expected number of text line groups to be recognized. - **`ReferenceGroupName`**: References another group for relative positioning or filtering. - **`SubGroups`**: Defines sub-groups within the text line specification for hierarchical organization. - **`TextLinesCount`**: Specifies the expected number of text lines to be processed. #### Output Control - **`OutputResults`**: Set to 1 to include this text line's results in the final output, or 0 to process internally without outputting. #### Character Correction - **`ConfusableCharactersCorrection`**: Configuration for correcting commonly confused characters (e.g., 0 vs O, 1 vs I).