Articles
A review of current technologies for detecting and deterring harmful birds in horticultural crops
Article number
1433_34
Pages
265 – 272
Language
English
Abstract
Humanity needs food to survive, and most food comes from crops.
Agricultural crops around the world are often destroyed by various pests, including birds.
They attack different crops, usually in large flocks, often causing major damage to farmers.
Specialists tried different methods of repelling birds from cultures (scarecrows, balloons, kites, models with raptor birds, reflective tapes, devices that emit sounds and ultrasound, lasers, etc., even drones equipped with various auditory or visual systems, and other), but none of these offer complete crop protection.
Recent research in the field, based on artificial intelligence, show new models of crop protection systems using machine learning algorithms, which help recognizing birds.
Bird detection and localization is a superior application of computer vision.
After the target bird has been identified, the system automatically selects the appropriate bird-rejection sound and plays it to specifically scare the birds away, emission of specific sounds and or ultrasound being different for each species.
Learning-based methods help detect and locate birds through images, this approach taking into consideration the conditions in which the birds are presented, with different shapes and sizes as also complex environments in which they survive.
In this regard, this paper presents a review of the current existing systems for the detection of harmful birds in crops, developed and recently presented by researchers.
Agricultural crops around the world are often destroyed by various pests, including birds.
They attack different crops, usually in large flocks, often causing major damage to farmers.
Specialists tried different methods of repelling birds from cultures (scarecrows, balloons, kites, models with raptor birds, reflective tapes, devices that emit sounds and ultrasound, lasers, etc., even drones equipped with various auditory or visual systems, and other), but none of these offer complete crop protection.
Recent research in the field, based on artificial intelligence, show new models of crop protection systems using machine learning algorithms, which help recognizing birds.
Bird detection and localization is a superior application of computer vision.
After the target bird has been identified, the system automatically selects the appropriate bird-rejection sound and plays it to specifically scare the birds away, emission of specific sounds and or ultrasound being different for each species.
Learning-based methods help detect and locate birds through images, this approach taking into consideration the conditions in which the birds are presented, with different shapes and sizes as also complex environments in which they survive.
In this regard, this paper presents a review of the current existing systems for the detection of harmful birds in crops, developed and recently presented by researchers.
Authors
A. Pruteanu, M.G. Matache, D.A. Carstea, R.D. Cristea
Keywords
detecting, images, harmful birds, crops, convolutional neural networks
Online Articles (41)