GARBAGE SEGREGATION USING IMAGE PROCESSING смотреть онлайн

Garbage is now being dumped into landfills without separation or separated improperly and this has led to a major problem in waste disposal over the past decade. This decreases efficiency in the process and also contributes more to pollution. The proper disposal of waste largely impacts the environment and public health. This helps reduce hazard emissions, thereby mitigating climate change.
A model is proposed to automatically separate garbage into recyclable and non-recyclable sections by using image processing algorithms and a robotic arm to physically separate the waste into separate bins.
Previous works in this topic have included methods that have tackled the challenge by using proximity sensors or using image processing algorithms that only focused on a single object at a time which would be inefficient and time consuming.
This mode processes an image of the garbage in front of the robotic arm and simultaneously identifies the recyclable objects at once using a Convolutional Neural Network (CNN) called You Only Look Once (YOLO), then systematically disposes of the waste into the respective bin. The model achieved a Mean Average Precision (mAP) of 61.49% with a maximum detection time of 300 milliseconds.

Acknowledgement
The project is sponsored by Karnataka State Council for Science and Technology (KSCST) under 43rd Series of Student Project Program and JSSATEB.


DR. NAVEEN N C - [email protected]
Ms. JYOTHSNA P - [email protected]
Ms. CAUVERY A - [email protected]
Mr. BHARGAV HEGDE - [email protected]
Ms. AKSHITHA Y V - [email protected]