Happy Earth Day! Today millions of people around the world will be discussing emerging green technologies and environmental innovation–we wanted to get in on the conversation too! We’ve written in the past about computer vision based wildlife conservation efforts but for this post we thought we would pick up a topic that’s a little less fluffy and arguably more smelly: computer vision and waste management.
Computer Vision and Waste Detection
Waste management might not be the first thing you think about when someone mentions computer vision but it’s actually an excellent use case for the technology because identifying trash is, at its most basic, a form of object detection. A slightly more complex example would be object classification, where a computer not only recognizes a piece of trash as a soda bottle but as a Coca-Cola bottle or as polyethylene terephthalate plastic, or PET as it is commonly referred to.
A computer vision model identifies different types of waste. Courtesy of GreyParrot’s website.
The Global Waste Crisis
Annually, human beings produce 2.01 billion tonnes of waste (that’s over 4 trillion pounds) and according to the World Bank. In high income countries like the United States and Europe almost 51% of all waste is capable of being recycled. In a study published in 2018 by the EPA, the agency reported that of the 292.4 million tons of municipal solid waste produced in the US, only 69 million tons was recycled, coming in around 24% of the total waste–which means roughly 76 million tons of recyclable material were not correctly processed and most likely ended up in a landfill.
How Can Computer Vision Help?
With the overseas demand for plastic recyclables diminishing the United States and other countries will have to look to domestic recycling options to process more of their waste. In order for this option to be viable the existing waste processing facilities need to be more efficient at sorting out recyclable materials. The current waste processing system in the US is a combination of locally regulated collection streams (different color trash bins designating different types of waste) and material recovery centers where waste is sorted by hand and weight on conveyor belts.
Building Waste Detection Datasets and Models
Computer vision, deep learning, and robotics offer a possible avenue for improving this process. The creation of waste material datasets like Recycleye’s WasteNet, which houses 2.5 million training images, is the first step to creating accurate computer vision models to detect different kinds of waste. The company has amassed several different datasets including Brand-Level and Material-Level recognition. Another power player in the sector is GreyParrot whose vision system can recognize over 40 categories of waste. We would love to know what annotation tool the companies are using.
WasteNet’s Database and Sub-Databases. Image courtesy of Recycleye’s Website.
Material Recovery centers are not the only area where waste detection models can be useful. CSIRO, Australia’s Natural Science Agency, has been experimenting with computer vision detection in local waterways. Waterways like canals and rivers are notorious litter traps and oftentimes feed into the ocean. CSIRO’s waste detection models are used to help “identify litter hot spots, implement better waste-related policies and improve waste management methods to make them safer, smarter and relatively cheaper.” The models were created from a dataset of over 14,500 individual items and 30 categories of classification.
Tools for Solutions
Global waste is projected to grow to 6.8 trillion pounds produced annually or 3.4 billion tonnes by 2050. It’s awesome to know that computer vision, deep learning, and AI can be counted as tools in the fight to clean up our planet. It is also clear that great computer vision solutions start with robust datasets and accurate models. Sixgill’s Sense AI Platform is the perfect tool for creating custom computer vision solutions, offering a low-code platform that allows even nontechnical users to annotate, train, and deploy computer vision solutions quickly and accurately. Solving the waste crisis is going to take a lot of brain power and innovation, but there are tools available in the fight.