Riverscape

Inspiration

Our project, Riverscape, draws inspiration from the captivating story of Eugenie Clark, known as the Shark Lady. Eugenie's unwavering passion for sharks and her pioneering contributions to the field of marine science serve as a powerful reminder of the profound impact individuals can have on our understanding and appreciation of the natural world.

Like Eugenie Clark, we believe that sharks are not to be feared, but rather admired and respected as essential members of marine ecosystems. Her relentless pursuit of knowledge and her efforts to educate the world about these graceful creatures have left an indelible mark on the scientific community and society as a whole.

Eugenie Clark's story serves as a beacon of inspiration, reminding us that women can break barriers and excel in traditionally male-dominated fields. Her journey showcases the strength, determination, and resilience required to challenge societal norms and pave the way for future generations of scientists.

By combining Eugenie Clark's legacy with our goal of addressing plastic waste in rivers, we are driven to create a sustainable future where the impact of human activities on marine life is minimized. Through the application of AI and technology, we aim to emulate Eugenie's unwavering dedication to understanding and preserving the ocean's delicate ecosystems.

Riverscape embodies the spirit of the Shark Lady, harnessing the power of AI, machine learning, and environmental consciousness to reduce plastic waste, protect our rivers, and create a sustainable environment where marine life can thrive. Inspired by Eugenie Clark's fearless pursuit of knowledge, we strive to make a positive impact, educating and inspiring others to join us in safeguarding our oceans for generations to come.

What it does

Riverscape is an innovative application that utilizes AI and machine learning to detect and track plastic waste in rivers. It leverages computer vision algorithms to analyze images and identify instances of plastic pollution. The application also incorporates geotagging to map the location of plastic waste, enabling users to visualize and understand the extent of the problem.

How we built it

Through the combination of web scraping with Beautiful Soup, geotagging, AI model training, and application deployment, we constructed Riverscape as an innovative solution to address plastic waste in rivers. By leveraging technology and data-driven approaches, we aim to contribute to the preservation of river ecosystems, the reduction of plastic pollution, and the promotion of sustainability.

To gather data for Riverscape, we utilized the Beautiful Soup library in Python. With Beautiful Soup, we performed web scraping on various sources, including websites, forums, and social media platforms, to collect relevant information and data on plastic waste in rivers. We extracted data such as locations, descriptions, and images of plastic pollution incidents, ensuring a comprehensive dataset for our project. Once the data was collected, we performed preprocessing tasks to clean and structure the data. We removed any irrelevant or duplicated entries, standardized the data format, and organized it for further analysis. Additionally, we manually labeled the images in the dataset, classifying them as either containing plastic waste or not, to facilitate supervised learning for our machine learning models.

Geotagging plays a crucial role in understanding the distribution and spatial patterns of plastic waste in rivers. Using the geolocation data associated with the collected images, we employed geotagging techniques to assign latitude and longitude coordinates to each plastic waste incident. This allowed us to map the locations of plastic pollution hotspots and gain insights into the spatial patterns and trends of plastic waste in rivers. With the preprocessed and labeled dataset, we trained our AI model using the Detectron2 library, which is built on the PyTorch framework. We fine-tuned the model using transfer learning techniques, leveraging pre-trained weights and architectures to improve the model's ability to detect and classify plastic waste in river images. We evaluated the performance of the model using metrics such as precision, recall, and accuracy to ensure its reliability and effectiveness.

Challenges we ran into

Throughout the development process, we faced various challenges. Obtaining a diverse and high-quality dataset for training the model was a significant hurdle. We had to gather and label a large number of images to ensure the model's effectiveness in different scenarios. Additionally, optimizing the model's performance to achieve real-time detection was another challenge we encountered. We had to experiment with different architectures, hyperparameters, and optimization techniques to strike the right balance between accuracy and speed.

Accomplishments that we're proud of

We are proud to have developed a functioning prototype of Riverscape that can accurately detect and track plastic waste in rivers. We achieved our goal of integrating AI and machine learning algorithms to address a critical environmental issue. Our application can assist in identifying areas of high plastic pollution, raising awareness among communities, and enabling targeted cleanup efforts. We are also proud of the scalability of our solution, as it can be adapted to other water bodies and contribute to a broader sustainability initiative.

What we learned

Through this project,we learned how to use beautifulsoup library for webscraping.we also learnt a few other technologies like geotagging.Throughout the development of Riverscape, we gained valuable insights into the challenges and complexities of plastic waste detection using AI. We learned about the importance of data quality and diversity in training accurate models. We also deepened our understanding of computer vision techniques, model optimization, and real-time inference. Additionally, we realized the significance of collaboration and interdisciplinary approaches in tackling complex problems like plastic pollution.

What's next for Riverscape

In the future, we envision expanding the capabilities of Riverscape by incorporating more advanced AI algorithms. We plan to enhance the model's ability to classify different types of plastic waste, such as bottles, bags, or microplastics. Furthermore, we aim to integrate real-time monitoring systems, such as drones or underwater robots, to collect live data on plastic waste. Our goal is to make Riverscape a comprehensive tool for environmental monitoring and decision-making, empowering communities and organizations to take effective actions in reducing plastic pollution in rivers and preserving our precious ecosystems..

Built With

• beautiful-soup
• detectron
• python
• pytorch