Struct RawImage

pub struct RawImage {

    pub make: String,
    pub model: String,
    pub clean_make: String,
    pub clean_model: String,
    pub width: usize,
    pub height: usize,
    pub cpp: usize,
    pub wb_coeffs: [f32; 4],
    pub whitelevels: [u16; 4],
    pub blacklevels: [u16; 4],
    pub xyz_to_cam: [[f32; 3]; 4],
    pub cfa: CFA,
    pub crops: [usize; 4],
    pub blackareas: Vec<(u64, u64, u64, u64)>,
    pub orientation: Orientation,
    pub data: RawImageData,
}

All the data needed to process this raw image, including the image data itself as well as all the needed metadata

Fields

make: String

camera make as encoded in the file

model: String

camera model as encoded in the file

clean_make: String

make cleaned up to be consistent and short

clean_model: String

model cleaned up to be consistent and short

width: usize

width of the full image

height: usize

height of the full image

cpp: usize

number of components per pixel (1 for bayer, 3 for RGB images)

wb_coeffs: [f32; 4]

whitebalance coefficients encoded in the file in RGBE order

whitelevels: [u16; 4]

image whitelevels in RGBE order

blacklevels: [u16; 4]

image blacklevels in RGBE order

xyz_to_cam: [[f32; 3]; 4]

matrix to convert XYZ to camera RGBE

cfa: CFA

color filter array

crops: [usize; 4]

how much to crop the image to get all the usable area, order is top, right, bottom, left

blackareas: Vec<(u64, u64, u64, u64)>

Areas of the sensor that is masked to prevent it from receiving light. Used to calculate black levels and noise. Each tuple represents a masked rectangle’s top, right, bottom, left

orientation: Orientation

orientation of the image as indicated by the image metadata

data: RawImageData

image data itself, has width*height*cpp elements

Implementations

impl RawImage

pub fn cam_to_xyz(&self) -> [[f32; 4]; 3]

Outputs the inverted matrix that converts pixels in the camera colorspace into XYZ components.

pub fn cam_to_xyz_normalized(&self) -> [[f32; 4]; 3]

Outputs the inverted matrix that converts pixels in the camera colorspace into XYZ components normalized to be easily used to convert to Lab or a RGB output space

pub fn neutralwb(&self) -> [f32; 4]

Not all cameras encode a whitebalance so in those cases just using a 6500K neutral one is a good compromise

pub fn normalized_pseudoinverse(inm: [[f32; 3]; 4]) -> [[f32; 4]; 3]

Normalize matrix so that inm * (1,1,1) is (1,1,1,1) and then invert

pub fn pseudoinverse(inm: [[f32; 3]; 4]) -> [[f32; 4]; 3]

Matrix inversion that deals with 4x3 matrices

pub fn cropped_cfa(&self) -> CFA

Returns the CFA pattern after the crop has been applied (and thus the pattern potentially shifted)

pub fn is_monochrome(&self) -> bool

Checks if the image is monochrome

Trait Implementations

impl Clone for RawImage

fn clone(&self) -> RawImage

Returns a copy of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for RawImage

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

Formats the value using the given formatter.