International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 05 | May 2023
p-ISSN: 2395-0072
www.irjet.net
Endangered Bird Species Classification Using Machine Learning Techniques Suhas Reddy B R 1, P V Bhaskar Reddy2, Vikramadhitya P S 3, Veluri Raviram Nikhil 4, Abhilash C 5 School of Computer Science and Engineering REVA University, Bengaluru, India --------------------------------------------------------------------------***----------------------------------------------------------------------have become a critical priority for researchers, Abstract - Birds are a diverse class of warm-blooded conservationists, and policymakers worldwide.
creatures, with around 10,000 living species presenting a range of characteristics and appearances. Although though individuals frequently enjoy viewing birds, accurate bird species identification requires an understanding of the field of ornithology. To address this issue, we offer a CNN-based automated model that can distinguish between several bird species using a test dataset. Our model was trained using a dataset of 7,637 pictures representing 20 distinct bird species, of which 1,853 were selected for testing. The deep neural network's design was developed to analyse the images and draw out traits for categorization. We tested a variety of hyperparameters and techniques, such data augmentation, to improve performance. According to our findings, the suggested model evaluated on the dataset had a promising accuracy of 98%. Our study also emphasises the value of utilising technology to safeguard and maintain endangered bird populations as well as the promise of convolutional neural networks for bird species identification. In summary, the suggested methodology can help with bird population identification and tracking, which will ultimately help with their preservation and protection. The model's accuracy may be increased, and its application can be broadened to cover other bird species.
Keywords:
Bird species, Machine Convolutional Neural Networks, Ornithology.
I.
One of the challenges in protecting endangered bird species is the ability to accurately identify and classify them. Birds can be challenging to identify due to their diverse appearances, behaviours, and songs. Accurate identification is crucial for conservation efforts as it enables researchers to track populations, monitor habitats, and design effective conservation strategies. Traditional methods of bird identification rely on visual observations and expert knowledge, which can be timeconsuming, labour-intensive, and error-prone. Additionally, the availability of experts in the field of ornithology is limited, making the task of bird identification even more challenging. To address these challenges, researchers have turned to machine learning techniques for automated bird species identification. Machine learning algorithms are capable of learning from large datasets and identifying patterns that humans may miss, making them an attractive solution for bird identification. In recent years, there has been a growing interest in applying machine learning to bird identification, with promising results. In addition to the importance of bird conservation, it is also important to note the potential impact of technological advancements in this field. With the rise of machine learning and artificial intelligence, it has become possible to use these tools to aid in conservation efforts. By automating the process of identifying endangered bird species, we can more efficiently and accurately track their populations and assess the success of conservation strategies.
Learning,
INTRODUCTION
The world is home to a diverse range of living creatures, each with unique characteristics and traits that make them fascinating to study and observe. Among these creatures, birds have captured the attention of humans for centuries, with their beautiful plumage, intricate behaviors, and important ecological roles. However, despite the fascination and admiration that birds evoke in us, many bird species are facing serious threats to their survival. Human activities such as deforestation, climate change, and pollution are causing the loss of habitats and food sources for birds, leading to declines in populations and even extinctions. In this context, the conservation and protection of endangered bird species
© 2023, IRJET
|
Impact Factor value: 8.226
In this research, we offer a paradigm for automatically classifying endangered birds. Our approach involves pre-processing bird images to extract features, and then training a machine learning model to classify the species. We explore the use of various machine learning algorithms, including deep convolutional neural networks, and evaluate their performance on a dataset of bird images from several
|
ISO 9001:2008 Certified Journal
|
Page 81