README.md

GDSC Solution Challenge - BeadyEyes

AI repo

How to run my code on local environment by using python

---

Text Recognition based on Hand Landmark (OCR)

What is it?

• Read the text indicated by your finger
• You can OCR recognize and retrieve the letter closest to the top of the part you are pointing with your finger.

Reference

• google-cloud-vision-OCR
• MediaPipe-Hand landmarks detection

Enviroment

• python version : 3.11

How to do?

$ git clone https://github.com/GDSC-Solution-Challenge-Team-4/BeadyEye-AI.git && cd BeadyEye-AI/OCR

$ pip install -r requirements.txt

• You need to have a Google Cloud Platform (GCP) key prepared.
• (Google Cloud Platform (GCP) key 가 준비되어 있어야 합니다.)

• Create GCP key

• After downloading the gcp key, change the file name to “ourAccountKey.json” and place it in the OCR folder.

• (gcp key를 다운받은 후 파일이름을 "ourAccountKey.json"으로 변경한 후 OCR 폴더에 넣어주세요)

You can get the results returned through the following code.

$ python -c "from handLandmark import text_pointer_uri; result = text_pointer_uri('{your_image_uri}');print('\nresult: ', result)"

You can OCR recognize and retrieve the letter closest to the top of the part you are pointing with your finger.

---

Example

Example1

• excute code

• result

  

Example2

• excute code

• result