Test neuroglancer with prediction output of Object Classification Workflow

[m-novikov]

We need to check how well neuroglancer handlers indexed color. We are interested in object classification output that is single-channel image with object index as value.
Supposedly neuroglancer segmentation layer does the job.
TODO:

• Test dataset uploaded to n5test
• Neuroglancer link demonstrating this dataset

[Tomaz-Vieira]

Here's a sample output from object classification, as run by cloud ilastik:

swift download -p out_Object-Predictions_0fb1c52a-1236-459e-9f07-e2dc4935f730.n5 n5test

I should add that the repeating tiled image is not a bug. I just copy-pasted the same input image multiple times to make a bigger one