Skip to main content

ImageProcessorTrait

edgefirst_image

Trait ImageProcessorTrait 

Source 
pub trait ImageProcessorTrait {
    // Required methods
    fn convert(
        &mut self,
        src: &TensorImage,
        dst: &mut TensorImage,
        rotation: Rotation,
        flip: Flip,
        crop: Crop,
    ) -> Result<()>;
    fn convert_ref(
        &mut self,
        src: &TensorImage,
        dst: &mut TensorImageRef<'_>,
        rotation: Rotation,
        flip: Flip,
        crop: Crop,
    ) -> Result<()>;
    fn render_to_image(
        &mut self,
        dst: &mut TensorImage,
        detect: &[DetectBox],
        segmentation: &[Segmentation],
    ) -> Result<()>;
    fn render_from_protos(
        &mut self,
        dst: &mut TensorImage,
        detect: &[DetectBox],
        proto_data: &ProtoData,
    ) -> Result<()>;
    fn render_masks_from_protos(
        &mut self,
        detect: &[DetectBox],
        proto_data: ProtoData,
        output_width: usize,
        output_height: usize,
    ) -> Result<Vec<MaskResult>>;
    fn set_class_colors(&mut self, colors: &[[u8; 4]]) -> Result<()>;
}

Required Methods§

Source

fn convert( &mut self, src: &TensorImage, dst: &mut TensorImage, rotation: Rotation, flip: Flip, crop: Crop, ) -> Result<()>

Converts the source image to the destination image format and size. The image is cropped first, then flipped, then rotated

§Arguments
  • dst - The destination image to be converted to.
  • src - The source image to convert from.
  • rotation - The rotation to apply to the destination image.
  • flip - Flips the image
  • crop - An optional rectangle specifying the area to crop from the source image
§Returns

A Result indicating success or failure of the conversion.

Source

fn convert_ref( &mut self, src: &TensorImage, dst: &mut TensorImageRef<'_>, rotation: Rotation, flip: Flip, crop: Crop, ) -> Result<()>

Converts the source image to a borrowed destination tensor for zero-copy preprocessing.

This variant accepts a TensorImageRef as the destination, enabling direct writes into external buffers (e.g., model input tensors) without intermediate copies.

§Arguments
  • src - The source image to convert from.
  • dst - A borrowed tensor image wrapping the destination buffer.
  • rotation - The rotation to apply to the destination image.
  • flip - Flips the image
  • crop - An optional rectangle specifying the area to crop from the source image
§Returns

A Result indicating success or failure of the conversion.

Source

fn render_to_image( &mut self, dst: &mut TensorImage, detect: &[DetectBox], segmentation: &[Segmentation], ) -> Result<()>

Source

fn render_from_protos( &mut self, dst: &mut TensorImage, detect: &[DetectBox], proto_data: &ProtoData, ) -> Result<()>

Renders detection boxes and segmentation masks from raw prototype data.

For YOLO segmentation models, this avoids materializing intermediate Array3<u8> masks. The ProtoData contains mask coefficients and the prototype tensor; the renderer computes mask_coeff @ protos directly.

Phase 1 implementation materializes masks internally and delegates to existing render paths. Phase 2 will compute masks in the GPU shader.

Source

fn render_masks_from_protos( &mut self, detect: &[DetectBox], proto_data: ProtoData, output_width: usize, output_height: usize, ) -> Result<Vec<MaskResult>>

Renders per-instance grayscale masks from raw prototype data at full output resolution.

Each mask is rendered at the detection’s bounding-box region using sigmoid(mask_coeff @ protos) without thresholding, producing continuous [0,255] values suitable for soft IoU computation.

Returns one MaskResult per detection with the bbox-cropped pixels.

Source

fn set_class_colors(&mut self, colors: &[[u8; 4]]) -> Result<()>

Sets the colors used for rendering segmentation masks. Up to 17 colors can be set.

Implementors§