image2layout-computer-vision 0.1.3

image processing and stuff

image2layout_computer_vision

An image processing module for some computer vision tasks (public module for image2layout)

Package Page: pypi

Features:

1. Text Detection and Recognition (OCR)
2. Color extraction (background and main foreground)

Installations

Install with python/conda [Linux]

1. (Optional) Conda

curl https://repo.anaconda.com/archive/Anaconda3-2023.03-1-Linux-x86_64.sh -o ~/conda.sh
bash ~/conda.sh -b -f -p /opt/conda
rm ~/conda.sh
conda init --all --dry-run --verbose

conda create -n cv python=3.10 -y
conda activate cv

3. Python libraries (python>=3.8)

# python -m pip install 'torch>=2.0' torchvision torchaudio
python -m pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu

python -m pip install datasets transformers scikit-learn
python -m pip install --upgrade datasets transformers scikit-learn Pillow numpy pandas
python -m pip install paddleocr paddlepaddle

python -m pip install --upgrade image2layout-computer-vision

Install with docker

OUTDATED - update in progress

sudo docker build --tag cv -f Dockerfile_cpu .

sudo docker run -it -p 0.0.0.0:8000:8000 -p 0.0.0.0:8001:8001 -v $(pwd):/app cv bash

From inside container

cd deployment
conda activate cv
python api_serve.py -n CV -p 8000

Usage

1. Run this python code to pre-download model weights

from image2layout_computer_vision import model_dispatch_layout
model_dispatch_layout._load()

2. Recognize texts

import image2layout_computer_vision as icv

# 2 lists of dicts with keys [text, box, score]
data_merged, data_raw = icv.detect_text_full('path/to/image.png')

3. Extract colors

import image2layout_computer_vision as icv

color_bg, color_fg = icv.extract_colors('path/to/image.png')

Build

(for building and uploading this package)

python -m pip install --upgrade pip
python -m pip install --upgrade build twine "keyring<19.0"

rm -rf dist
python -m build
python -m twine upload dist/* --verbose