Spi-VSTL: Image data collection platform using off-the shelf hardware for vertically supported crops in state-of-the-art glasshouses
Horticulture crop growers are moving from conventional to protected crops, aiming for quality food production utilising fewer resources. Skilled labour for monitoring and maintaining crops in these compact environments has been identified as a major cost and can be reduced using automated image-base...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | HardwareX |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2468067225000021 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Horticulture crop growers are moving from conventional to protected crops, aiming for quality food production utilising fewer resources. Skilled labour for monitoring and maintaining crops in these compact environments has been identified as a major cost and can be reduced using automated image-based crop monitoring. There is a range of protected cropping infrastructures targeting different types of crops. Image data collection platforms are needed to be tailored according to the infrastructure and nature of the crop. Available research or commercial-purpose image data collection platforms for indoor crops are mostly targeted at movable and small plants compared to vertically supported tall plants. Customising existing commercial systems for this specific type of crop is costly. This paper proposes a low-cost image data collection platform for monitoring vertically supported tall crops in order to reduce labour costs while expanding the monitoring tasks for maintaining better crop growth. Off-the-shelf hardware and electronic components accessible from Australia are used for this development. The proposed platform runs manually on concrete flow and on pipe rail systems found in state-of-the-art commercial glasshouse settings. The proposed motorised platform has been tested with 30 kg, and speed was measured as an average minimum of 0.06 ms−1 and an average maximum of 0.47 ms−1. The usability of the proposed design has been proved with a published data set and research on plant height estimation. Other use cases and room for further development are also discussed. |
|---|---|
| ISSN: | 2468-0672 |