perf: faster cropping; parallel imread/crop; mt_loop and mp_loop helper functions

[yxlao]

Summary

This PR focuses on optimizing performance for cropping and reading operations by leveraging parallel processing. The addition of mt_loop and mp_loop functions introduces flexible utilities for applying any operation in parallel, offering significant speed improvements for processing large sets of images or data.

Changes

1. Performance Enhancements for Image Cropping:
   Improved the image cropping performance by introducing a new version of the ct.image.compute_cropping function for enhanced efficiency.

   import camtools as ct
   image = ct.io.imread("example.png")
   cropping = ct.image.compute_cropping(image)

2. Parallel Image Reading and Cropping:
   Integrated parallel processing for reading and cropping images, utilizing the newly added mt_loop and mp_loop functions in camtools/utility.py. These functions enable multi-threaded and multi-process execution, respectively, significantly speeding up the operations when working with multiple images.

   # This will be automatically parallelized
   ct crop-boarders --same_crop *.png

3. Utility Functions for Parallel Loops:
   Added mt_loop for multi-threading and mp_loop for multi-processing in camtools/utility.py, facilitating parallel execution of functions over iterables. These utility functions aim to generalize parallel execution patterns, making it easy to apply any function in a multi-threaded or multi-processed manner across a list of inputs.

   • Multi-threading Example:

     import camtools as ct
     inputs = range(10)
     results = ct.utility.mt_loop(my_function, inputs)

   • Multi-processing Example:

     import camtools as ct
     inputs = range(10)
     results = ct.utility.mp_loop(my_function, inputs)