TY - GEN
T1 - Measurement of sugarcane lodging extent using machine vision
AU - Momin, Abdul
AU - Grift, Tony
AU - Baier, James
N1 - This project was funded by the Energy Biosciences Institute, University of Illinois at Urbana-Champaign. Thanks to Mr. Brandon Gravois from Edgard, LA, USA for his support in conducting the field experiment.
PY - 2024
Y1 - 2024
N2 - Stalk lodging in sugarcane is common and ubiquitous, negatively affecting harvesting performance and overall productivity. Reliable methods for quantifying the extent of lodging, whether off-line or on-the-go, are currently non-existent. This study measured the extent of lodging based on images taken in the field during harvesting time, and subsequently, used the data to classify stalk lodging into erect, lodged and recumbent classes. The imagery was acquired using high-resolution color cameras (GO-PRO: HERO4), mounted at various locations on harvesting machinery. The sugarcane materials included erect as well as lodged sugarcane, both unburnt and burnt. The analysis comprised segmentation of images followed by vertical edge related pixel (verp) accumulation of images that were rotated in a range from 0 to 180ºwith a 1ºincrement. The analysis revealed that, in this limited study, approximately 42% of stalks fell into the recumbent class. The method as proposed may have potential for future development of an on-the-go sugarcane lodging sensor that could improve harvesting efficiency and overall productivity.
AB - Stalk lodging in sugarcane is common and ubiquitous, negatively affecting harvesting performance and overall productivity. Reliable methods for quantifying the extent of lodging, whether off-line or on-the-go, are currently non-existent. This study measured the extent of lodging based on images taken in the field during harvesting time, and subsequently, used the data to classify stalk lodging into erect, lodged and recumbent classes. The imagery was acquired using high-resolution color cameras (GO-PRO: HERO4), mounted at various locations on harvesting machinery. The sugarcane materials included erect as well as lodged sugarcane, both unburnt and burnt. The analysis comprised segmentation of images followed by vertical edge related pixel (verp) accumulation of images that were rotated in a range from 0 to 180ºwith a 1ºincrement. The analysis revealed that, in this limited study, approximately 42% of stalks fell into the recumbent class. The method as proposed may have potential for future development of an on-the-go sugarcane lodging sensor that could improve harvesting efficiency and overall productivity.
KW - Burnt cane
KW - edge detection
KW - erect
KW - recumbent
KW - unburnt cane
UR - http://www.scopus.com/inward/record.url?scp=85206091043&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85206091043&partnerID=8YFLogxK
U2 - 10.13031/aim.202400775
DO - 10.13031/aim.202400775
M3 - Conference contribution
AN - SCOPUS:85206091043
T3 - 2024 ASABE Annual International Meeting
BT - 2024 ASABE Annual International Meeting
PB - American Society of Agricultural and Biological Engineers
T2 - 2024 American Society of Agricultural and Biological Engineers Annual International Meeting, ASABE 2024
Y2 - 28 July 2024 through 31 July 2024
ER -