Bird species classification based on image using Convolutional Neural Network / Adam Izzat Azmi

For numerous people nowadays, determining the species of birds and classifying them is getting challenging. To reliably describe bird species without relying on human labour, research has been done in this area. To identify and categorise bird species using digital images of their forms, colours, an...

全面介紹

Saved in:
書目詳細資料
主要作者: Azmi, Adam Izzat
格式: Thesis
語言:English
出版: 2024
主題:
在線閱讀:https://ir.uitm.edu.my/id/eprint/95534/1/95534.pdf
標簽: 添加標簽
沒有標簽, 成為第一個標記此記錄!
實物特徵
總結:For numerous people nowadays, determining the species of birds and classifying them is getting challenging. To reliably describe bird species without relying on human labour, research has been done in this area. To identify and categorise bird species using digital images of their forms, colours, and patterns is the goal of this research. As part of the approach used in this project, a dataset of bird photos was gathered, the data was processed, and a Convolutional Neural Network model was trained to accurately identify and categorise the species of birds. The results of this study show the value of employing Convolutional Neural Network to identify birds because they successfully categorise birds in a variety of contexts with high accuracy rates. The actual work done includes data collecting from the Kaggle dataset, Convolutional Neural Network implementation, training the model, and performance evaluation. The acquired results demonstrate the potential of CNNs-based bird species categorization systems in raising interest in learning and increasing the success rate of monitoring bird populations. By offering fresh perspectives and approaches to the classification of bird species, this research advances the subject and creates new opportunities for global improvements in the study of animals. Finally, it is envisaged that the classification of bird species based on an image system will aid in expanding our understanding of and research into bird species, particularly in Malaysia.