haozihong/EvaDB-Google-Cloud-Vision-Ext

Google Cloud Vision API

Setup

Prepare a Google Cloud project ID and a Google Cloud API key.

(Optional) Create a new virtual environment and activate it:

python -m venv .env
source .env/bin/activate

Ensure that the local Python version is >= 3.8. Install the required libraries:

pip install -r requirements.txt

Run the Example

Open gcloud_obj_detect_example.py, fill in the Google Cloud API key and project ID, and then run it:

python gcloud_obj_detect_example.py

The annotated images will be saved in the same directory with annotated suffixes.

Or Try it in Jupyter Notebook

Install and start the Jupyter Lab:

pip install jupyterlab
jupyter lab

Open gcloud_obj_detect_example.ipynb, fill in the Google Cloud API key and project ID in the 5th cell, and run the cells one by one.

Example in SQL

CREATE OR REPLACE FUNCTION gvision_obj_detect
IMPL 'google_cloud_vision_object_detector.py';

DROP TABLE IF EXISTS MyImage;
LOAD IMAGE 'imgs/bicycle_example.png' INTO MyImage;
LOAD IMAGE 'imgs/example2.jpeg' INTO MyImage;
LOAD IMAGE 'imgs/example3.jpg' INTO MyImage;

SELECT gvision_obj_detect(data) from MyImage;