🌎 📡 SATELLITE-IMAGE-OBJECTS-RECOGNITION

📝 Description

• This implemantation is based on official Mask-rcnn Algorithm.
• It is Mask-rcnn based SATELLITE-IMAGE-OBJECTS-RECOGNITION model using Flask framework.
• Model is trained on 1200 images of classes Forest, River and Building and each class containing around 400 images.

🎥 Inference Video Link :-

• https://www.youtube.com/watch?v=1rMTWfE7GEA&t=38s

🎯 Inference demo

🎯 Inference Images

1)image1

• 

Output:-

• 

2)image 2

• 

Output:-

• 

💻 Requirements

Download this requirements.txt file & install.

• To install run :-

$ pip install -r requirements.txt

⏳ Dataset

• Download the dataset for custom training
• Dataset:- UC Merced Land Use Dataset
• http://weegee.vision.ucmerced.edu/datasets/landuse.html