Aizu University Technology: Wildlife Detection System: K23-003
Real-time detection of animals with high accuracy using machine learning.
In areas close to mountains, damage caused by wildlife to humans and crops is a persistent issue. In recent years, there have been a series of sightings even in urban areas, and there is a demand to reduce accidents caused by sudden encounters. To address this challenge, this research aims to develop a wildlife alert system using machine learning. This system automatically detects wildlife in place of humans, alerts people nearby with sound and light, and informs registered email users of detection information. The current alert system has the following features: 1. Detection using a learning model of wildlife generated from collected images with convolutional neural networks. 2. Detection of wildlife using a microcontroller board included in the alert device (a type of edge AI), completing the process from sensing to sound or light alert in about 3 seconds, and from sensing to email notification in about 1 minute. 3. Operates on solar panels and batteries, eliminating the need for commercial power sources.
- Company：Tohoku Techno Arch Co., Ltd.
- Price：Other
