lost-sia
Version:
Single Image Annotation Tool
26 lines (16 loc) • 1.51 kB
text/mdx
# Coordinate Systems
SIA uses a scene-graph-like scheme for handling annotations.
Outside of SIA, every coordinate is normalized; it is a pair of values between 0 and 1 that represents the fraction of the image.
For example, the coordinate `{x: 0.5, y: 1}` would be in the bottom line and in the middle row of the image.
Inside the `Canvas` component, a different system is used: Instead of using the fraction, the Canvas uses the pixel values of the so-called stage.
## Stage handling
The stage is a compromise between the image and the available space for the canvas.
It represents the biggest possible scaling of the image that fits onto the canvas and can be made without affecting the aspect ratio of the image.
To convert the coordinates from the image into the stage coordinate system, each coordinate can be multiplied with the `imageToStageFactor`.
## Translations
The manual translations (when the user moved the image using wasd or the mouse) and the zoom are handled using the svg transform.
Therefore its not needed to translate the coordinates according to these modifications.
However, mouse events are not in relation to the stage or image but to the current page.
Because of this, the mouse coordinates need to be translated from the page coordinate system into the stage coordinate system by countering the svg transformations and applying the `pageToStageOffset`.
