Android Application of Leaf Identification System

Android Application of Leaf Identification System

leaf identification

CNN

Android application

TensorFlow Lite

Leaf identification image is consistently a difficult task when using computer vision. The convolutional component extraction methods on images have their impediment and limitation, such as low accuracy, are not adaptable and less promising when converted to genuine application. The reasons are the lack of dataset needed to build a recognition model. Likewise, using the computer as a tool is bothering as it restricts the task in the research lab only. Convolutional Neural Network (CNN) shows a great solution for the computer version. Subsequently, this project utilizes the CNN’s properties to solve the image classification task, and the CNN model chosen is run in Phyton coding in TensorFlow Lite. It is similar to TensorFlow’s running code, but this project focused on building an Android application. It can perform faster and produce high accuracy results. There are four types of leaves involved in this project: betik, kari, pudina, and cengal. As a result, the model could reach around 99% accuracy with a 0.176 error rate. Ultimately, an Android application called Leaf identification is created. The model is sent and integrated into the apps that work with a concentrated information base to help put away and deal with the pictures. Hence, an Android leaf image identifier using CNN is proposed to solve the stated problem and is believed to contribute to education and research.

Minhat, Mohammad Hazwan Hakim

Saon, Sharifah

Mahamad, Abd Kadir

Umi Fadlillah

International Journal of Interactive Mobile Technologies (iJIM); Vol. 16 No. 15 (2022); pp. 62-77

1865-7923

International Association of Online Engineering (IAOE), Vienna, Austria

2022-08-17

Copyright (c) 2022 Hazwan Hakim, Sharifa, Dr Kadir , Umi Fadlillah

https://creativecommons.org/licenses/by/4.0

application/pdf

eng

info:eu-repo/semantics/article

info:eu-repo/semantics/publishedVersion

Peer-reviewed Article