The Fusion of CNN and MLP Algorithm as High-Performance Classification for Identification of Healthy and Unhealthy Leaves

Title

The Fusion of CNN and MLP Algorithm as High-Performance Classification for Identification of Healthy and Unhealthy Leaves

Creator

Gaba, Ira; Ramamurthy, B.

Description

In this study, it encapsulates the results of the work carried out with the Convolutional Neural Network, Multi-layer Perceptron, and hybrid of CNN and MLP classifier for the recognition of a tea leaf. The leaves are categorized into distinct classes by analysis and identification and recognition, which benefits both the buyer and the farmer by enabling the seller to sell tea leaves based on the quality of the leaf. Nowadays there is more advancement in the field of agriculture. But it is always the latest subject to study in the field of agriculture for the analysis and to identify quality of leaves. Many AI methodologies can be used for identification and recognition and further their fusion with different techniques or methods that can be applied to address the issue and to acquire the better accuracy. In this CNN, MLP and the hybrid of CNN-MLP are employed for determining the accuracy and this can further help in classifying the leaf in different grades like best quality, average quality, and worst quality, as for the future scope. Then the feature selection algorithms are implemented based on the different selection methods such as ANOVA, information gain, feature importance, and the random forest, which will reduce the number of parameters, at the end classification is carried out for identification of the leaf. 2026, Greater Mekong Subregion Academic and Research Network, Asian Institute of Technology. All rights reserved.

Source

GMSARN International Journal;Volume;20;Issue;1;pp.49-59

Date

01-01-2026

Publisher

Greater Mekong Subregion Academic and Research Network, Asian Institute of Technology

Subject

Convolutional Neural Network; Feature extraction and selection; methods; Multi-layer Perceptron; Pre-processing

Coverage

Gaba I., Computer Science Department, Christ University, Bangalore, 560067, India; Ramamurthy B., Computer Science Department, Christ University, Bangalore, 560067, India

Rights

Restricted Access; Hardcopy may be available in the library

Relation

ISSN: 19059094;

Format

online

Language

English

Type

Article

Identifier

https://www.scopus.com/pages/publications/105011671317?origin=resultslist