Preliminary evaluation of real-time sensing of harvester losses by machine vision
Paper
Paper/Presentation Title | Preliminary evaluation of real-time sensing of harvester losses by machine vision |
---|---|
Presentation Type | Paper |
Authors | McCarthy Chery |
Journal or Proceedings Title | Proceedings of the 43rd Annual Conference of the Australian Society of Sugar Cane Technologists (ASSCT 2022) |
Journal Citation | 43, pp. 108-110 |
Page Range | 108-110 |
Number of Pages | 3 |
Year | 2022 |
Publisher | Australian Society of Sugar Cane Technologists |
Place of Publication | Australia |
ISBN | 9781713859215 |
Web Address (URL) of Conference Proceedings | https://www.proceedings.com/65361.html |
Conference/Event | 43rd Annual Conference of the Australian Society of Sugar Cane Technologists (ASSCT 2022) |
Event Details | 43rd Annual Conference of the Australian Society of Sugar Cane Technologists (ASSCT 2022) Parent Australian Society of Sugar Cane Technologists Conference Delivery Online Event Date 19 to end of 22 Apr 2022 Event Location Mackay, Australia Event Venue Mackay Entertainment and Convention Centre |
Abstract | Sugar losses during cane cleaning in mechanical harvesters are estimated to cause millions of dollars of lost income per year. Existing commercially available loss-monitoring devices do not directly sense losses from the material expelled during cane cleaning in the harvester. Development of technologies for real-time, accurate and consistent measurement of harvester losses is required to achieve improved efficiency of harvesting with reduced losses. A proof-of-concept machine-vision sensor containing cameras with visible light and non-visible light sensitivity has been developed for the purpose of real-time sensing of harvester losses. Initial trials were conducted in October 2020 in the Gordonvale region. Primary extractor fan speed was varied for the trials, and a Sugar Research Australia field team recorded losses data using the Infield Sucrose Loss Measurement System. The trials enabled machine-vision sensor data to be compared with sugar expelled from the harvester under a range of field conditions. Machine-vision analysis has indicated a coefficient of determination of between 0.72 and 0.93 for prediction of sugar loss from image data from a combination of camera sensors. Further analysis is presently being undertaken on trial data collected in 2021 under different field conditions. Machine vision has potential to detect sugar losses for the purpose of providing real-time feedback to harvester operators. Ultimately, such a sensor has potential use to automatically detect and provide recommendations for harvester settings in real-time to minimise losses. |
Keywords | automation; billets; harvest losses; Image analysis; infield sucrose loss measurement system |
Public Notes | Files associated with this item cannot be displayed due to copyright restrictions. |
Funder | Sugar Research Australia |
Byline Affiliations | University of Southern Queensland |
https://research.usq.edu.au/item/yyyv7/preliminary-evaluation-of-real-time-sensing-of-harvester-losses-by-machine-vision
46
total views0
total downloads3
views this month0
downloads this month