delta.data

Click on the signatures below to get a complete description.

Module for data manipulations and input/output operations.

@author: jblugagne

Functions

binarizerange(array)	This function will binarize a numpy array by thresholding it in the middle of its range
data_augmentation(images_input, aug_par[, ...])	Data augmentation function
estimateClassweights(gene[, num_samples])	Estimate the class weights to use with the weighted categorical cross-entropy based on the output of the trainGenerator_track output.
estimate_seg2D_classweights(mask_path[, ...])	Estimate the weights to assign each class in the weight maps
getfnfromprototype(prototype, fileorder[, ...])	Generate full filename for specific frame based on file path, prototype, fileorder, and filenamesindexing
histogram_voodoo(image[, num_control_points])	This function kindly provided by Daniel Eaton from the Paulsson lab.
illumination_voodoo(image[, num_control_points])	This function inspired by the one above.
kernel(n)	Get kernel for morphological operations
postprocess(images[, square_size, min_size, ...])	A generic binary image cleaning function based on mathematical morphology.
predictCompilefromseg_track(img_path, ...[, ...])	Compile an inputs array for tracking prediction with the tracking U-Net, directly from U-Net segmentation masks saved to disk.
predictGenerator_seg(files_path[, ...])	Get a generator for predicting segmentation on new image files once the segmentation U-Net has been trained.
readreshape(filename[, target_size, ...])	Read image from disk and format it
saveResult_seg(save_path, npyfile[, ...])	Saves an array of segmentation output images to disk
saveResult_track(save_path, npyfile[, ...])	Save tracking output masks to disk
seg_weights(mask[, classweights, w0, sigma])	This function computes the weight map as described in the original U-Net paper to force the model to learn borders (Slow, best to run this offline before training)
seg_weights_2D(mask[, classweights])	Compute custom weightmaps designed for bacterial images where borders are difficult to distinguish
shift(image, vector[, order])	Image shifting function
tracking_weights(track, segall[, halo_distance])	Compute weights for tracking training sets
trainGenerator_seg(batch_size, img_path, ...)	Generator for training the segmentation U-Net.
trainGenerator_track(batch_size, img_path, ...)	Generator for training the tracking U-Net.
zoomshift(I, zoomlevel, shiftX, shiftY[, order])	This function zooms and shifts images.