This project deals with image processing. It mainly revolves around generating captions for the uploaded image, detecting edges and corners with features and find anomalies based on comparison. Captioning is done using CNN and LSTM. Edges are detected using Canny Edge Detection algorithm, corners are found using Harris Corner Algorithm and anomalies are found using absolute pixel difference and thresholding.
CNN and LSTM Model for captioning.
Python - 3.10
To download captioning dataset, visit
Flickr_8K Dataset
Flickr_8K Text
Before installing modules, create an environment. To create it run the command
python -m venv .env
Enter environment
cd .env
To clone repository, run
git clone https://github.com/jeyprabu/Image-Decoding-and-Captioning.git
Then,
cd Image-Decoding-and-Captioning
To install the modules, run the following command.
pip install -r requirements.txt
Once installed, run the app
flask run
or python app.py
jeyprabu / image-decoding-and-captioning Goto Github PK
View Code? Open in Web Editor NEWA web application developed to generate captions from images. It can also detect edges and corners of an image. Furthermore, it can perform comparative anomaly detection.