A two-stage model for enhanced mango leaf disease detection using an innovative handcrafted spatial feature extraction method and knowledge distillation process

The economic stability of many countries is closely tied to agriculture, where crop quality and yield are heavily affected by weather conditions and disease control. Unpredictable climate patterns and plant diseases remain major challenges, emphasizing the need for more reliable disease detection me...

Full description

Saved in:
Bibliographic Details
Main Authors: Mohammad Manzurul Islam, Mst. Nasrat Jahan Niva, Abdullahi Chowdhury, Saleh Masum, Rifat Ara Shams, Taskeed Jabid, Md. Sawkat Ali, Md. Mostofa Kamal Rasel, Muhammad Firoz Mridha
Format: Article
Language:English
Published: Elsevier 2025-11-01
Series:Ecological Informatics
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1574954125003747
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849222584859623424
author Mohammad Manzurul Islam
Mst. Nasrat Jahan Niva
Abdullahi Chowdhury
Saleh Masum
Rifat Ara Shams
Taskeed Jabid
Md. Sawkat Ali
Md. Mostofa Kamal Rasel
Muhammad Firoz Mridha
author_facet Mohammad Manzurul Islam
Mst. Nasrat Jahan Niva
Abdullahi Chowdhury
Saleh Masum
Rifat Ara Shams
Taskeed Jabid
Md. Sawkat Ali
Md. Mostofa Kamal Rasel
Muhammad Firoz Mridha
author_sort Mohammad Manzurul Islam
collection DOAJ
description The economic stability of many countries is closely tied to agriculture, where crop quality and yield are heavily affected by weather conditions and disease control. Unpredictable climate patterns and plant diseases remain major challenges, emphasizing the need for more reliable disease detection methods. Traditionally, plant disease identification has relied on visual examination, a method that is often inaccurate. To address this, our study proposes a two-stage model for improved disease detection in mango leaves. In the first stage, we implement an innovative, block-based feature extraction technique using Local Directional Patterns (LDP) and Local Directional Pattern variance (LDPv) on a comprehensive dataset, MangoLeafBD, consisting of 4000 images, achieving satisfactory results in terms of detection accuracy, sensitivity, specificity, and false negative rate. In the second stage, we introduce a Knowledge Distillation (KD) process to further enhance model performance by transferring knowledge from a larger teacher model to a smaller student model. Our results demonstrate significant advancement, with the KD-enhanced model achieving an improvement in detection accuracy from 89.2% to 95.6%, sensitivity from 7.8% to 4.1%, and specificity from 71.2% to 88.9% for Anthracnose disease. Similar improvements were observed in detecting other diseases, making our approach a more robust and efficient solution for mango plant disease detection.
format Article
id doaj-art-c5d49e73444247f29049abbf8528f364
institution Kabale University
issn 1574-9541
language English
publishDate 2025-11-01
publisher Elsevier
record_format Article
series Ecological Informatics
spelling doaj-art-c5d49e73444247f29049abbf8528f3642025-08-26T04:14:11ZengElsevierEcological Informatics1574-95412025-11-019110336510.1016/j.ecoinf.2025.103365A two-stage model for enhanced mango leaf disease detection using an innovative handcrafted spatial feature extraction method and knowledge distillation processMohammad Manzurul Islam0Mst. Nasrat Jahan Niva1Abdullahi Chowdhury2Saleh Masum3Rifat Ara Shams4Taskeed Jabid5Md. Sawkat Ali6Md. Mostofa Kamal Rasel7Muhammad Firoz Mridha8Department of Computer Science and Engineering, East West University, Aftabnagar, Dhaka 1212, Bangladesh; Corresponding author.Department of Computer Science and Engineering, East West University, Aftabnagar, Dhaka 1212, BangladeshDepartment of Computer Science and Engineering, East West University, Aftabnagar, Dhaka 1212, BangladeshDepartment of Information and Communication Engineering, University of Rajshahi, Charukala Road, Rajshahi 6205, BangladeshDepartment of Information and Communication Engineering, University of Rajshahi, Charukala Road, Rajshahi 6205, BangladeshDepartment of Computer Science and Engineering, East West University, Aftabnagar, Dhaka 1212, BangladeshDepartment of Computer Science and Engineering, East West University, Aftabnagar, Dhaka 1212, BangladeshDepartment of Computer Science and Engineering, East West University, Aftabnagar, Dhaka 1212, BangladeshDepartment of Computer Science, American International University Bangladesh, Khilkhet, Dhaka 1229, BangladeshThe economic stability of many countries is closely tied to agriculture, where crop quality and yield are heavily affected by weather conditions and disease control. Unpredictable climate patterns and plant diseases remain major challenges, emphasizing the need for more reliable disease detection methods. Traditionally, plant disease identification has relied on visual examination, a method that is often inaccurate. To address this, our study proposes a two-stage model for improved disease detection in mango leaves. In the first stage, we implement an innovative, block-based feature extraction technique using Local Directional Patterns (LDP) and Local Directional Pattern variance (LDPv) on a comprehensive dataset, MangoLeafBD, consisting of 4000 images, achieving satisfactory results in terms of detection accuracy, sensitivity, specificity, and false negative rate. In the second stage, we introduce a Knowledge Distillation (KD) process to further enhance model performance by transferring knowledge from a larger teacher model to a smaller student model. Our results demonstrate significant advancement, with the KD-enhanced model achieving an improvement in detection accuracy from 89.2% to 95.6%, sensitivity from 7.8% to 4.1%, and specificity from 71.2% to 88.9% for Anthracnose disease. Similar improvements were observed in detecting other diseases, making our approach a more robust and efficient solution for mango plant disease detection.http://www.sciencedirect.com/science/article/pii/S1574954125003747Detecting diseasesMango leavesMachine learningSpatio-temporal features
spellingShingle Mohammad Manzurul Islam
Mst. Nasrat Jahan Niva
Abdullahi Chowdhury
Saleh Masum
Rifat Ara Shams
Taskeed Jabid
Md. Sawkat Ali
Md. Mostofa Kamal Rasel
Muhammad Firoz Mridha
A two-stage model for enhanced mango leaf disease detection using an innovative handcrafted spatial feature extraction method and knowledge distillation process
Ecological Informatics
Detecting diseases
Mango leaves
Machine learning
Spatio-temporal features
title A two-stage model for enhanced mango leaf disease detection using an innovative handcrafted spatial feature extraction method and knowledge distillation process
title_full A two-stage model for enhanced mango leaf disease detection using an innovative handcrafted spatial feature extraction method and knowledge distillation process
title_fullStr A two-stage model for enhanced mango leaf disease detection using an innovative handcrafted spatial feature extraction method and knowledge distillation process
title_full_unstemmed A two-stage model for enhanced mango leaf disease detection using an innovative handcrafted spatial feature extraction method and knowledge distillation process
title_short A two-stage model for enhanced mango leaf disease detection using an innovative handcrafted spatial feature extraction method and knowledge distillation process
title_sort two stage model for enhanced mango leaf disease detection using an innovative handcrafted spatial feature extraction method and knowledge distillation process
topic Detecting diseases
Mango leaves
Machine learning
Spatio-temporal features
url http://www.sciencedirect.com/science/article/pii/S1574954125003747
work_keys_str_mv AT mohammadmanzurulislam atwostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT mstnasratjahanniva atwostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT abdullahichowdhury atwostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT salehmasum atwostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT rifatarashams atwostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT taskeedjabid atwostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT mdsawkatali atwostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT mdmostofakamalrasel atwostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT muhammadfirozmridha atwostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT mohammadmanzurulislam twostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT mstnasratjahanniva twostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT abdullahichowdhury twostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT salehmasum twostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT rifatarashams twostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT taskeedjabid twostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT mdsawkatali twostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT mdmostofakamalrasel twostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess
AT muhammadfirozmridha twostagemodelforenhancedmangoleafdiseasedetectionusinganinnovativehandcraftedspatialfeatureextractionmethodandknowledgedistillationprocess