简体中文 · English

This repository contains some simple modules for Crack Detection and Semantic Segmentation, mainly using Window 11 + Conda virtual environment.

The environment configuration is here.

/crack_detection_util
    /image_util.py
        class MyImage
        class ChunkAndMerge
    /evaluate_util.py
        class Evaluator

Refer to usage for package usage.

The environment configuration of modules is listed below, including system environment and python module dependencies.

Generally, even when the versions of referred python modules are different, the program is able to run normally.

Python version: 3.10.13
OS: Windows

opencv-contrib-python-headless==4.9.0.80
opencv-python-headless==4.9.0.80

numpy==1.23.5
pandas==2.2.1
Pillow==9.5.0 (important)

torch==1.13.1+cu116
torch-tb-profiler==0.4.3
torchaudio==0.13.1+cu116
torchvision==0.14.1+cu116

tensorboard==2.15.2
tensorboard-data-server==0.7.2
tensorflow==2.15.0
tensorflow-estimator==2.15.0
tensorflow-intel==2.15.0
tensorflow-io-gcs-filesystem==0.31.0

Using a higher version of the Pillow module may cause errors in the use of the pycocotools module. It is recommended to install 9.5.0 version.

Check this file if any critical module is missing.