Image library

This Module provide factory and utility for imaging processing

Refer to API: neuralib.imglib

Array Factory

This module defines the ImageArrayWrapper, a subclass of numpy.ndarray that wraps image data and provides chainable image processing methods. It allows you to process images fluently using a series of method calls, and then display or further analyze the results as a standard NumPy array.

Example of Chainable Processing

from neuralib.imglib.array import image_array
import matplotlib.pyplot as plt

# Load an image (from a file path or a NumPy array)
img = image_array("path/to/image.jpg")

# Process the image: convert to grayscale, apply Gaussian blur, then perform edge detection.
processed = img.to_gray().gaussian_blur(ksize=[5, 5], sigma_x=2.0, sigma_y=2.0).canny_filter()

# Display the processed image using matplotlib.
plt.imshow(processed, cmap='gray')
plt.title("Processed Image")
plt.show()

Method Reference

Method	Description
`to_gray()`	Converts the image to grayscale.
`flipud()`	Flips the image vertically (upside down).
`fliplr()`	Flips the image horizontally (left-to-right).
`select_channel(channel)`	Extracts a specified color channel (e.g. ‘r’, ‘g’, or ‘b’) as a grayscale image.
`view_2d(flipud=False)`	Converts a multi-channel image to a 2D representation (grayscale) with an option to flip vertically.
`gaussian_blur(ksize, sigma_x, sigma_y)`	Applies a Gaussian blur to the image using the specified kernel size and sigma values.
`canny_filter(threshold_1, threshold_2)`	Applies the Canny edge detection algorithm on the grayscale version of the image.
`binarize(thresh, maxval)`	Converts the image to a binary image using thresholding.
`denoise(h, temp_win_size, search_win_size)`	Denoises the image using non-local means denoising.
`enhance_contrast()`	Enhances the image contrast via histogram equalization.
`local_maxima(channel)`	Computes the local maxima on a specified color channel after channel extraction.

CV2 Labeller

Simple CV2-based viewer/labeller GUI for image sequences

Use Cases:

viewing the image sequences
label each image and save as csv dataframe (human-eval for population neurons activity profile)

Load sequences from a directory

See Option

$ python -m neuralib.imglib.labeller -h

Example

$ python -m neuralib.imglib.labeller -D <DIR>

API call

from neuralib.imglib.labeller import SequenceLabeller

directory = ...
labeller = SequenceLabeller.load_from_dir(directory)
labeller.main()

Load sequences from sequences array

from neuralib.imglib.labeller import SequenceLabeller

arr = ...   # numpy array with (F, H, W, <3>)
labeller = SequenceLabeller.load_sequences(arr)
labeller.main()