A research group led by Kim Sun Il, a researcher at the Faculty of Metal Engineering, has introduced his invention, an intelligent laser bird repeller with a deep neural network.
By comparison with existing repellers, the system of this bird repeller, whose design and database are based on sufficient experimental research into birds in various kinds and sizes, is programmed to undergo constant update, thus ensuring greater efficiency in bird detection and classification.
It detects birds by means of a camera, decides the mode of attack suitable for the detection results and sends signals to the laser driver device before attacking.
It consists of three parts ― a laser attack unit combined with an outdoor monitoring camera, a laser driver unit and a computer processing unit.
The laser attack unit consisting of the bunch of green (532nm), blue (420nm) and red (650nm) lasers of 1W is embedded in the outdoor monitoring camera. Once an image is sent to the computer, it detects and tracks the bird and sends a signal to the bird repeller device using artifical intelligence technology. This device emanates laser light driver signals according to the kind and size of birds and the intensity of light dependent on the weather change.
The interface consists of monitor, statistics, log, main setting, etc.
With 98.7% detection rate and 95.3% repelling rate, the fully automated system is applicable to all places including farms, fish farms, orchards and airfields that are exposed to the damage by birds.
© 2021 KumChaek University of Technology