Date of Publication
4-10-2024
Document Type
Master's Thesis
Degree Name
Bachelor of Science in Mechanical Engineering (Honors) - Ladderized
College
Gokongwei College of Engineering
Department/Unit
Mechanical Engineering
Thesis Advisor
Dr. Laurence A. Gan Lim
Defense Panel Chair
Eng. Jeremias A. Gonzaga
Defense Panel Member
Dr. Ronnie Concepcion
Eng. Conrad Allan Jay R. Pantua
Abstract/Summary
This thesis addresses the critical issue of rural poverty and stalled agricultural advancement in the Philippines by proposing a modernization initiative in agricultural practices, particularly in the post-harvest handling of tomatoes. Recognizing the deficiencies in conventional harvesting and sorting techniques which often lead to significant economic losses and decreased produce quality, this study aims to develop and evaluate a computer vision-based sorting system designed to differentiate between ripe and unripe tomatoes while minimizing mechanical damage to the fruit. The research leverages the YOLOv8 architecture to create an advanced vision system equipped with an infrared camera, facilitating accurate classification and detection under varying lighting conditions. A unique aspect of this approach is the use of a water flume design, innovatively conceived to reduce mechanical strain during sorting. This system setup proved effective, achieving a high object detection confidence score of 0.875 and a classification accuracy rate of 95% during trials. Such performance not only underscores the practical viability of the system but also highlights its potential to replace traditional sorting methods, thereby enhancing efficiency and reducing waste. This thesis contributes to the field by filling a significant gap in research related to post-harvest technologies that specifically mitigate mechanical damage through the immediate implementation of sorting systems post-harvest. The findings suggest that integrating sophisticated machine learning models and specially designed mechanical setups can profoundly impact agricultural productivity, offering a sustainable solution to the challenges hindering economic growth in rural agricultural communities. The implementation of such technology could greatly improve sectoral performance, enhance economic outcomes for farmers, and support the broader agenda of modernizing agriculture in the Philippines.
Abstract Format
html
Language
English
Format
Electronic
Keywords
Crops—Postharvest technology
Recommended Citation
De Leon, J. L. (2024). Design of post-harvest automated water conveyor classification apparatus using an infrared camera with convolution neural networks for kamatis (Solanum lycopersicum). Retrieved from https://animorepository.dlsu.edu.ph/etdm_mecheng/27
Upload Full Text
wf_yes
Embargo Period
4-10-2025