ROI coordinates

The ROIData class is responsible for converting napari labels layers into bounding box coordinates that can be used to add grids in SBEMimage. These ROIs are stored as MaskROI objects and include the bounding box coordinates relative to the centre of the overview image. ROIData.add_masks finds the bounding boxes for each labeled object in the labels layer.

Because the labels layer can have any 2D affine transform, these coordinates are transformed using the layer's translation, affine_matrix and scale properties.

To add a grid into SBEMimage, the TCPServer.add_grid method is used with the ROI ID (labels layer integer), ROI centre coordinates relative to the centre of the overview layer, ROI width and height, and coordinates of the overview (using the metadata stored in LiveViewer).

Individual grid tiles can be activated depending on the shape of the object using the 3D mask stored in MaskROI.mask. This mask is obtained by applying the above transformations to the original binary mask. Grid tiles can be activated using the TCPServer.update_grid_tiles_with_mask using the ROI ID and the current 2D slice from the 3D mask. The 2D slice is obtained with the MaskROI.get_current_slice and the current z coordinate from SBEMimage.