How kayakers can guide conservation by monitoring with GoPros and Machine Learning

Описание: Purpose Paddling proposes a groundbreaking yet highly accessible method for kayakers to monitor the seafloor using simple action cameras like GoPros. We are working to design a streamlined device that attaches a camera directly under the kayak. Crucially, this setup connects to a continuous GPS receiver above the water, automatically mapping and geo-tagging your underwater video footage as you paddle. To add necessary scientific rigor, the device will incorporate three parallel lasers, projecting a visual reference onto the seafloor to accurately estimate the size of marine life and benthic structures.

The real power of this proposal lies in what happens after you return to the shore. We are developing a specialized Python machine learning script designed to analyze this underwater footage automatically. Instead of researchers spending countless hours manually reviewing video, our script will rapidly identify benthic species, perform presence and absence analysis of vital seaweed and kelp forests, and automatically count sea urchins to monitor destructive overgrazing. This technology transforms raw, accessible GoPro footage into precise, location-based ecological data.

I imagine a world where every single recreational kayak tour brings back valuable information to help save our oceans. To make this a reality, I need your help to build a centralized digital platform where paddlers from around the globe can easily upload their geo-tagged videos. This platform will run the machine learning analysis and immediately share the insights with conservationists, marine biologists, and policymakers. We want to ensure that this continuous stream of crowdsourced data does not just sit on a server, but is used directly to trigger and guide proactive restoration of our coastal waters.