Plastic Detection

Inspiration

My inspiration is to reduce the trash like plastic which pollutes the soil and water.

What it does

It detect the plastic items like covers, bottles.

How we built it

The program built in two stages. The first section detects waste by returning areas with lots of white light through binary thresholding, and areas with a high concentration of edges. Plastic, glass, and metal are generally reflective whereas organic matter tends to not be. By thresholding with a high color value.

Challenges we ran into

The Challenges are to change the input file, change the variables in the script globals.py, which is a script that only conatins the path for the input file. Other images are abvailbe in the folder called testimages. arious stages of the threshold mask can be saved and shown in the reduce.py script. They are commented out. It is recommended that this script is run on its own without the bash script if you want to save/see the stages of the binary mask. reduce.py also contains the parameters for the image processing.

Accomplishments that we're proud of

The program automatically saves the classified and unclassified images with bounding boxes.

What we learned

I have learned how detect a image by doing this project.

What's next for Plastic Detection

I have tried to detect the trash to reduce the usage of plastic. For future work we going to detect the trash in water and quantity of water polluted.

Built With

• python
• tensorflow