in

#Advanced wild bird repellent system using deep-learning and lasers. #BirdRepellent

Automatic wild bird repellent system that is based on deep-learning-based wild bird detection and integrated with a laser rotation mechanism

The content discusses the model selection, image data collection, and annotation process for a wild bird detection system using a mask R-CNN deep learning method. The model is trained on a dataset of wild bird images collected using a camera with motion detection. The images are manually selected and annotated using VGG Image Annotator software. The trained model is then evaluated for wild bird detection performance. The system architecture includes a wild bird detection unit, computing unit, and laser control unit for automatic bird repellent. Different laser scanning strategies are evaluated for their effectiveness in repelling wild birds.

Field experiments were conducted at a duck farm to test the proposed system. The experimental setup involved installing the system near feed buckets and recording the number of wild birds using cameras. The results were quantified using a mask R-CNN model to count the number of wild birds in each image captured. Two indicators, daily bird repulsion rate and hourly bird repulsion rate, were used to assess the effectiveness of the system. Statistical analysis was performed on the experimental results to determine the impact of the system on wild bird numbers. A negative binomial regression model was used to analyze the data and assess the system’s efficacy in repelling wild birds.

Source link

Source link: https://www.nature.com/articles/s41598-024-66920-2

What do you think?

Leave a Reply

GIPHY App Key not set. Please check settings

The Future of AI Development: Orchestrating a Symphony of Intelligence | by Quentin Drummond Anderson | Jul, 2024

Orchestrating a Symphony of Intelligence: The Future of AI #AIInnovation

Boost Your Efficiency with These Top Cutting-Edge AI Tools

Top AI tools to enhance efficiency in your work #efficiencyboost