Home / Current Issue / Paper 1705306
Fish Species Recognition using Convolutional Neural Networks for Biodiversity Monitoring
Subject area: Science,Engineering and Technology · Area of research: Machine Learning, Deep Learning
Abstract
In this research, our goal is to develop an automated system using Convolutional Neural Networks (CNNs) for recognizing fish species, thereby enhancing biodiversity monitoring efficiency. We focus on three specific fish species?Brachaluteres jacksonianus, Cantherhines dumerilii, and an additional species?and begin by curating a diverse dataset of high-resolution images from various aquatic environments. Utilizing the Python Imaging Library (PIL) and other open-source libraries, we preprocess, explore, and visually represent the dataset. The essence of our study lies in training a CNN model to accurately classify fish species based on their unique visual features. The model undergoes meticulous training and validation with a carefully divided dataset, emphasizing high accuracy and generalization. Visualizations of training history, encompassing accuracy and loss metrics, enable comprehensive evaluation across multiple epochs. Our proposed CNN model shows promise in revolutionizing biodiversity monitoring, offering a scalable and automated solution for precise fish species recognition. Successful implementation could streamline data collection, enhance ecological study efficiency, and provide critical insights for sustainable aquatic ecosystem management, marking a significant stride in integrating technology into conservation practices.
Keywords
Fish species recognition, Convolutional Neural Networks (CNN), Biodiversity monitoring, Automated classification, Ecological sustainability.
References
[1] Goodwin, A., Padmanabhan, S., Hira, S. et al. Mosquito species identification using convolutional neural networks with a multitiered ensemble model for novel species detection.1st July 2021.
[2] Dagher and D. Barbara, ‘‘Facial age estimation using pre-trained CNN and transfer learning,’’ Multimedia Tools Appl., vol. 80, pp. 20369–20380, Mar. 2021.
[3] L. Miao, W. Jingxian, L. Hualong, H. Zelin, Y. XuanJiang, H. Xiaoping, Z. Weihui, Z. Jian, and F. Sisi, ‘‘Method for identifying crop disease based on CNN and transfer learning,’’ Smart Agriculture., vol. 1, no. 3, pp. 46–55, 2019.
[4] Darwish, D. Ezzat, and A. E. Hassanien, ‘‘an optimised model based on convolutional neural networks and orthogonal learning particle swarm optimization algorithm for plant diseases diagnosis,’’ Swarm Evol. Compute. vol. 52, Feb. 2020, Art. No. 100616.
[5] D. Song, ‘‘Classification of spiders,’’ Sichuan J. Zool., vol. 2, pp. 37–41, Mar. 1985.
[6] M. L. Lim, M. F. Land, and D. Li, ‘‘Sex-specific UV and fluorescence signals in jumping spiders,’’ Science, vol. 315, no. 5811, p. 481, 2007.
[7] QIANJUN CHEN, YONGCHANG DING, CHANG LIU, JIE LIU, AND TINGTING HE. “Research on spider sex recognition from images based on deep learning”. August 30, 2021.
[8] MadsDyrmann,HenrikKarstoft,Henrik SkovMidtiby, “Plant species classification using deep convolutional neural network”.September 13,2016.
[9] Dhruv Rathi, Sushant Jain, Dr. S. Indu. “Underwater fish species classification using CNN and deep learning”. December 30, 2018.
[10] Zhijian Zhou, Meng Zhang, Jiefuchen, Xuqingwu. (2020). Detection and classification of multi-magnetic targets using Mask-RCNN (2020).
How to cite this paper
@article{1705306,
author = {Amit Kumar Pandey, Dr. Santosh Singh, Ankush Sushil Singh, Shravan Shivanand Kamat},
title = {Fish Species Recognition using Convolutional Neural Networks for Biodiversity Monitoring},
journal = {Iconic Research And Engineering Journals},
year = {2023},
volume = {7},
number = {6},
pages = {223-227},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1705306.pdf},
abstract = {In this research, our goal is to develop an automated system using Convolutional Neural Networks (CNNs) for recognizing fish species, thereby enhancing biodiversity monitoring efficiency. We focus on three specific fish species?Brachaluteres jacksonianus, Cantherhines dumerilii, and an additional species?and begin by curating a diverse dataset of high-resolution images from various aquatic environments. Utilizing the Python Imaging Library (PIL) and other open-source libraries, we preprocess, explore, and visually represent the dataset. The essence of our study lies in training a CNN model to accurately classify fish species based on their unique visual features. The model undergoes meticulous training and validation with a carefully divided dataset, emphasizing high accuracy and generalization. Visualizations of training history, encompassing accuracy and loss metrics, enable comprehensive evaluation across multiple epochs. Our proposed CNN model shows promise in revolutionizing biodiversity monitoring, offering a scalable and automated solution for precise fish species recognition. Successful implementation could streamline data collection, enhance ecological study efficiency, and provide critical insights for sustainable aquatic ecosystem management, marking a significant stride in integrating technology into conservation practices.},
keywords = {Fish species recognition, Convolutional Neural Networks (CNN), Biodiversity monitoring, Automated classification, Ecological sustainability.},
month = {December},
}