แจ้งเอกสารไม่ครบถ้วน, ไม่ตรงกับชื่อเรื่อง หรือมีข้อผิดพลาดเกี่ยวกับเอกสาร ติดต่อที่นี่ ==>
หากไม่มีอีเมลผู้รับให้กรอก thailis-noc@uni.net.th
หากไม่มีอีเมลผู้รับให้กรอก thailis-noc@uni.net.th
Parkpoom Chaisiriprasert. Deep learning-based ripeness classification of mangosteen under natural imaging conditions. (). King Mongkut's University of Technology North Bangkok. Central Library. : , 2025.
Deep learning-based ripeness classification of mangosteen under natural imaging conditions
Name: Mintra Boondach
Organization :
Rangsit University. College of Digital Innovation Technology
Email :
mintra.b68@rsu.ac.th
Name: Parkpoom Chaisiriprasert
Organization :
Rangsit University. College of Digital Innovation Technology
Email :
parkpoom.c@rsu.ac.th
LCSH:
Image processing
-- Digital techniques
LCSH:
Agriculture
-- Technological innovations
LCSH:
Fruit
-- Quality control
Abstract:
Mangosteen is one of Thailands most valuable tropical fruits, with ripeness levels significantly influencing its taste, texture, market value, and export quality. Conventional ripeness assessment relies on manual visual inspection, which is subjective and unsuitable for large-scale operations. To address this limitation, this study proposes an automated image-based ripeness classification system. A dataset of 2,478 mangosteen images, categorized into five ripeness stages and captured under varied natural conditions, was used for training and evaluation. The methodology begins with preprocessing steps, including color balance and Retinex, hue-guided enhancement, and CLAHE combined with sharpening, to improve image quality and emphasize key ripeness features. Feature extraction is then performed using a CSPDarknet backbone, a PAN-FPN neck, and a decoupled detection head to capture both low- and high-level visual cues. The extracted features are subsequently evaluated with three models: YOLOv12, Faster R-CNN, and EfficientNetV2-B0. Experimental results show that YOLOv12 achieved the best performance with an average mAP@0.5 of 74.0%, surpassing Faster R-CNN (41.0%) and EfficientNetV2-B0 (52.0%), confirming the effectiveness of the proposed system as a robust, non-invasive solution for smart agriculture and post-harvest quality control.
King Mongkut's University of Technology North Bangkok. Central Library
Address:
BANGKOK
Email:
library@kmutnb.ac.th
Created:
2025
Modified:
2026-03-04
Issued:
2026-03-04
บทความ/Article
application/pdf
BibliograpyCitation :
In Prince of Songkla University, Phuket Campus. College of Computing. The 9th International Conference on Information Technology (InCIT 2025) (pp.414-419). Phuket : Prince of Songkla University
eng
©copyrights King Mongkut's University of Technology North Bangkok
RightsAccess:
ใช้เวลา
0.034853 วินาที
Mintra Boondach
| Title | Contributor | Type |
|---|---|---|
| Deep learning-based ripeness classification of mangosteen under natural imaging conditions
มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าพระนครเหนือ Mintra Boondach;Parkpoom Chaisiriprasert | บทความ/Article |
Parkpoom Chaisiriprasert
| Title | Contributor | Type |
|---|---|---|
| Deep learning-based ripeness classification of mangosteen under natural imaging conditions
มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าพระนครเหนือ Mintra Boondach;Parkpoom Chaisiriprasert | บทความ/Article | |
| Multi-branch vision transformer for early skin cancer detection
มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าพระนครเหนือ Jirachat Rujiwan;Khachonkit Chuiad;Parkpoom Chaisiriprasert | บทความ/Article | |
| Enhanced ResNet-50 for concrete damage classification in structural health monitoring
มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าพระนครเหนือ Teerath Thesniyom;Surachet Khanman;Apicha Deearom;Parkpoom Chaisiriprasert | บทความ/Article | |
| An explainable vision transformer-based ensemble framework for automated tire wear assessment
มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าพระนครเหนือ Puttita Asavapurk;Khachonkit Chuiad;Parkpoom Chaisiriprasert | บทความ/Article | |
| Comparative study of hedonic motivation, shopping lifestyle, and trust on impulse buying in social commerce platforms
มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าพระนครเหนือ Kanyapa Phromsing;Nitthanet Natthakunlanan;Parkpoom Chaisiriprasert | บทความ/Article |