Fume Analysis is a key check used by mining engineers while blasting. It focuses on checking specific fumes which are toxic in nature, This is a image processing / computer vision approach to filter and find percentage of these toxic fumes based on colors
This is a standard usecase and can be used to filter any moving object in a stabilised camera with consistent camera position.
Following is a series of snap showing the three phases (Raw | Background Subtraction | Harmful Fume Subtraction)
- Python3.7
- If Docker Approach is selcted:
- Docker
- Docker Compose
The program can run in either of the two following ways:
-
Running the python file directly by passing the video url with the python file:
pip install -r requirements.txt
python fume_detect_extract_classify.py --vid {vid_path}
-
I've also made it as a complete dockerized server with proper authentication. Following is the way to run the dockerized version:
cd dockerized_flask_container
sudo docker-compose build
sudo docker-compose up
Resource Chart for Rest API:
Resources | URL | Method | Param | Status | Param Body type |
---|---|---|---|---|---|
Register | /register | post | uname, pass | 200 OK, 301 username already exist | JSON |
Classify | /classify | post | uname, pass, video_file | 200 OK, 301, 302 Incorrect id or pass, 303 Out of token, 304 Invalid input | Form |
Database.ini:
I have used a public database but you can replace it with your credentials in the web/req_files/database.ini file