implement the cell diameter parameter

[marius10p]

The default Cellpose models are optimized for an average cell diameter of 30 pixels. Results can be really bad if the cell diameter is significantly larger or smaller than this.

[yxdragon]

did you mean the image should be zoomed to make sure the object 30 diameter? then input the models?

def get_flow(nets, img, cn=[0,0], sample=1, size=512, tile=True, work=1, callback=progress)

• nets: the nets loaded upon.

• img: the image to process

• cn: the cytoplasm and nucleus channels

• sample: if not 1, we scale it. (only avalible when tile==True)

• size: when tile==True, this is the tile size, when tile==False, we scale the image to size.

• tile: if True, method try to process image in tiles. else resize the image.

• work: open multi-thread to process the image. (GPU not recommend)

now some parameter added.

in tile mode: the input image would scale by sample factor, then use the size as tile.

in no tile mode: sample disabled, the input image would be scale to size x size.