Development of Pest detection system using Convolutional neural network in Mung Bean production

dc.contributor.authorHarold P. Ejes and Rogelio Joshua C. Miner III
dc.date.accessioned2026-02-20T00:35:57Z
dc.date.available2026-02-20T00:35:57Z
dc.date.issued2024
dc.description.abstractMung bean (Vigna radiata L.) is a nutritious legume traditionally cultivated in the Philippines, but its production faces challenges like low yields, pest infestations and limited market opportunities. The study uses Convolutional Neural Networks (CNN) to develop a system for develop a system for detecting pests in mung beans, evaluate CNN's performance in pest detection, and compare the accuracy of pictures taken with 8MP and 12MP cameras. The system is coded in Visual Studio using YOLOv5 and Python. It is trained with pictures of pest professionally taken by the photographers of BPI and tested with pictures of the pest from the field where pest infested naturally. Accuracy id determined using precision, recall and F1 score. Additionally, the accuracy comparison between 8MP and 12MP camera images is analyzed using a confusion matrix and T-test as its statistical tools. The result shows the overall accuracy of the developed system and is recorded at 98%, indicating that the model performs well in general. The accuracy is justified by the model's result of Precision, Recall and F-1 Score, which are 1.00, 1.00 and 0.98 respectively. A confusion matrix resulted in 97.67% and 98.83% for Android and Iphone images, respectively. Moreover, the T-test showed no significant difference, with a 0.61924 T-value and a critical T-value 2.05. This implies that the model maintains high accuracy regardless of the phone used for capturing pest images.
dc.identifier.urihttp://granarium.clsu.edu.ph/handle/123456789/962
dc.language.isoen_US
dc.relation.supervisorNICASIO C. SALVADOR, M.Sc.
dc.titleDevelopment of Pest detection system using Convolutional neural network in Mung Bean production
dc.typeThesis
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
H.P. EJES, R.J.C. MINER III.pdf
Size:
4.24 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
3.12 KB
Format:
Item-specific license agreed to upon submission
Description:
Collections