Abstract:
One of the biggest challenges facing tourists visiting Rwanda's national parks is the frustration of not being able to see their preferred animals. This can lead to disappointment and a negative overall experience for tourists, as well as challenges for park management. In this study, we aim to explore the potential of using machine learning, embedded computing system, and a client application to improve the animal-viewing experience for tourists at Rwanda's national parks.
To address this problem, we developed a system that combines a machine learning model (You Only Look Once, or YoloV5) with a Raspberry Pi and a client application (developed in React JS) to classify images and show the location of animals. This system utilizes cameras that are already deployed across the park, specifically in key areas where they are used to monitor the safety of the animals and ensure that tourists and guides are following the rules. The cameras capture videos in the form of frames, and the resulting images are processed by the machine-learning model on the Raspberry Pi to identify and classify the animals present in each image. The client application, developed in React JS, allows tourists to view the location of the animals on a map of the park.
In this research, we present the results of a one-component system and demonstrate its potential to improve the animal-viewing experience for tourists. It is important to note that to address this issue, the system needs to be deployed to multiple locations in the park, following the areas where the cameras are already installed. With the help of the guides and security personnel, we can identify additional locations to deploy the system to. With further resources and time, this approach could be scaled to cover the entire park, enabling tourists to plan their visits and easily locate their preferred animals
Overall, our research suggests that the use of machine learning and a client application has the potential to greatly enhance the animal-viewing experience for tourists at Rwanda's national parks, and could provide valuable insights for park management.
