Dataset requirements

Hard requirements

Requirements which have to be met in order for ML training and segmentation to run.

Homogeneous bitdepth

Bitdepth is the same for all training and subsequent segmentation images.

Homogeneous channel number

Number of channels is the same for all training and subsequent segmentation images. The alpha channel in png files is automatically removed and not used for training and segmentation.

Minimum image and dataset size

For training the minimum image size is 128*128 pixel (recommended: 1024*1024 pixel) per plane and the minimum total number of pixels are 724*724 pixel.

Maximum image size

Currently, all data processing is done in-memory, which limits the sizes of images that can be processed. The following are limits for a set of image parameters. The real limits for your application might differ depending on image sizes, dimensions and properties.
  • Maximum processable 2D plane size for images without Time or Z dimension
Image format
Image size (pixels)
RGB 8 bit
16 000 x 16 000
Grayscale 16 bit
17 000 x 17 000
  • Maximum number of planes along the Time or Z dimension for stack of images
Stack of RGB 8 bit images
Image size (pixels)
Number of planes
512 x 512
1024 x 1024
2646 x 2056
Stack of grayscale 16 bit images
Image size (pixels)
Number of planes
512 x 512
1024 x 1024
2464 x 2056

Best practices

Requirements which should be met to produce optimal results.

Same acquisition mode

Images are acquired in the same acquisition mode for both training and segmentation (e.g. both times with reflection mode, not training with reflected image and segmentation with transmitted image).

Size of objects/representative region

  • Objects or representative regions should be of similar pixel size across training and segmentation images
  • For optimal results, object sizes should be no larger than 512*512 pixel

Image size

  • The smallest image axis across all training images determines the size of the area that the trained model "sees". Thus, image sizes should be homogeneous and not vary drastically in size to avoid excluding valuable image context. (Does not apply if all training images are above 1024*1024 pixel.)
  • For segmentation, the minimum recommended image size is 1024*1024 pixel or the smallest image axis across all training images. The threshold is set by which ever quantity is smaller.
If you have any further questions about dataset requirements please reach out to us at [email protected]
Last modified 1mo ago