Struct DBModel 

pub struct DBModel { /* private fields */ }

Pure DB model implementation.

This model implements the core DB architecture and can be configured for different detection tasks through preprocessing and postprocessing configs.

Implementations

impl DBModel

pub fn new( inference: OrtInfer, resizer: DetResizeForTest, normalizer: NormalizeImage, postprocessor: DBPostProcess, ) -> DBModel

Creates a new DB model.

pub fn preprocess( &self, images: Vec<ImageBuffer<Rgb<u8>, Vec<u8>>>, ) -> Result<(ArrayBase<OwnedRepr<f32>, Dim<[usize; 4]>>, Vec<ImageScaleInfo>), OCRError>

Preprocesses images for detection.

pub fn infer( &self, batch_tensor: &ArrayBase<OwnedRepr<f32>, Dim<[usize; 4]>>, ) -> Result<ArrayBase<OwnedRepr<f32>, Dim<[usize; 4]>>, OCRError>

Runs inference on the preprocessed batch.

pub fn postprocess( &self, predictions: &ArrayBase<OwnedRepr<f32>, Dim<[usize; 4]>>, img_shapes: Vec<ImageScaleInfo>, score_threshold: f32, box_threshold: f32, unclip_ratio: f32, ) -> DBModelOutput

Postprocesses model predictions to bounding boxes.

pub fn forward( &self, images: Vec<ImageBuffer<Rgb<u8>, Vec<u8>>>, score_threshold: f32, box_threshold: f32, unclip_ratio: f32, ) -> Result<DBModelOutput, OCRError>

Runs the complete forward pass: preprocess -> infer -> postprocess.

Trait Implementations

impl Debug for DBModel

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.