OCROptions<T>
Options & Configuration · AiDotNet.ComputerVision.OCR
Options for OCR models.
Known derived types: TrOCROptions<T>
Constructors
| Constructor | Summary |
|---|---|
OCROptions() |
Properties
| Property | Summary |
|---|---|
OCRMode Mode { get; set; } |
OCR mode (scene text, document, or both). |
T ConfidenceThreshold { get; set; } |
Minimum confidence threshold. |
TextDetectionModel DetectionModel { get; set; } |
Text detection model to use. |
TextRecognitionModel RecognitionModel { get; set; } |
Text recognition model to use. |
bool CorrectSkew { get; set; } |
Whether to correct skewed text. |
bool DetectOrientation { get; set; } |
Whether to detect text orientation. |
bool GroupTextLines { get; set; } |
Whether to group text into lines/paragraphs. |
bool UsePretrained { get; set; } |
Whether to use pretrained weights. |
int MaxRecognitionWidth { get; set; } |
Maximum input width for the recognition model. |
int MaxSequenceLength { get; set; } |
Maximum sequence length for recognition. |
int RecognitionHeight { get; set; } |
Input image height for the recognition model. |
int? RandomSeed { get; set; } |
Seed for weight initialization, so two models built from equal options start from equal weights. |
string CharacterSet { get; set; } |
Character set/vocabulary for recognition. |
string[] SupportedLanguages { get; set; } |
Supported languages. |