EXPERIMENTAL TOOLS

Here you can find a set of experimental tools; this code is provided as-is, with no support or warranty.

 

---

LIST DATASETS

This script lists all the datasets in an instance, allows to select one, and copies its dataset_id to the clipboard.

Requirements

Python 3.6 or higher installed; the necessary packages can be installed via pip:

pip install requests pyperclip

Usage

Run the script with the following command:

python3 list_datasets.py url api-key

Parameters

• url: the base URL of the API
• api_key: the API key for authentication

Licenses

• pyperclip: Distributed under the BSD 3-Clause License.
• requests: Distributed under the Apache License 2.0.

 

---

DATASET REPORT GENERATOR

This script generates a comprehensive report for a dataset. It fetches data from the MVI API, processes the images and labels, and creates a detailed report in HTML, PDF, and Excel formats.

Features

• Filters files based on date range or quantity
• Processes images and labels, creating annotated images
• Generates an Excel report with existing labels data
• Creates pie charts for label distribution
• Generates HTML and PDF reports with detailed information and statistics

Requirements

Python 3.6 or higher installed; the necessary packages can be installed via pip:

pip install argparse jinja2 matplotlib opencv-python-headless openpyxl pandas Pillow requests weasyprint

Usage

Run the script with the following command:

python3 dataset_report_generator.py url api_key dataset_id --files <number_of_files> --start-date "dd-MMM-yyyy" --end-date "dd-MMM-yyyy"

(an example of date format dd-MMM-yyyy is: 01-Jan-2000)

Parameters

• url: the base URL of the API
• api_key: the API key for authentication
• dataset_id: the ID of the dataset to be processed
• --files: (optional) the number of files to process
• --start-date: (optional) the start date for filtering files (format: dd-MMM-yyyy)
• --end-date: (optional) the end date for filtering files (format: dd-MMM-yyyy)
• if the optional parameters are not included, a report with the entire contents of the dataset will be generated.

Licenses

• argparse: Distributed under the Python Software Foundation License Version 2.
• cv2 (OpenCV): Distributed under the Apache License 2.0.
• matplotlib: Distributed under the Matplotlib License (BSD-style).
• opencv-python-headless: Distributed under the MIT License. master/LICENCE).
• pandas: Distributed under the BSD 3-Clause License.
• Pillow: Distributed under the Historical Permission Notice and Disclaimer (HPND).
• requests: Distributed under the Apache License 2.0.