fluent-cv

A fluent style interface intended for working with images using opencv. Many times you may want to apply a couple filters, tweak the settings and compare results. Keeping track of changes, adding preview, timing or saving can take a few seconds and lines of code that distracts from the process.

Fluent_cv provides a simple way to apply functions, preview and/or save the result and keep track of which parameters were used to generate which result, e.g.:

import cv2
from fluent_cv import ImageProcessor

processor = ImageProcessor().load_image("data/test.jpg")

(processor
        .apply_filter_function(cv2.normalize, None, alpha=0, beta=255, norm_type=cv2.NORM_MINMAX)
        .apply_filter_function(cv2.medianBlur, 5)
        .preview()
        .apply_filter_function(cv2.bilateralFilter, 11, 75, 75)
        .optional_save()  # save y/n, add tags if desired
)

Save results with steps and settings for quick comparison, e.g.:

test_edited_revision_1.json:

{
    "original_filepath_loaded": "data/test.jpg",
    "timestamp": "2024-12-15T11:30:21.025135",
    "processing_steps": [
        {
            "filter_name": "loaded image: data/test.jpg"
        },
        {
            "filter_time": 0.000023,
            "filter_name": "normalize",
            "parameters": {
                "args": [
                    "None"
                ],
                "kwargs": {
                    "alpha": "0",
                    "beta": "255",
                    "norm_type": "32"
                }
            }
        },
        {
            "filter_time": 0.001002,
            "filter_name": "medianBlur",
            "parameters": {
                "args": [
                    "5"
                ],
                "kwargs": {}
            }
        },
        {
            "filter_time": 0.002997,
            "filter_name": "bilateralFilter",
            "parameters": {
                "args": [
                    "11",
                    "75",
                    "75"
                ],
                "kwargs": {}
            }
        }
    ]
}