Abstract
Utilizing image data for yield estimation is a key topic in modern agriculture. This paper addresses the difficulty of counting wheat spikelets using images, to improve yield estimation in wheat fields. A wheat spikelet image dataset was constructed with images obtained by a smartphone, including wheat ears in the flowering, filling, and mature stages of reproduction. Furthermore, a modified lightweight object detection method, YOLOv5s-T, was incorporated. The experimental results show that the coefficient of determination (R2) between the predicted and true values of wheat spikelets was 0.97 for the flowering stage, 0.85 for the grain filling stage, and 0.78 for the mature stage. The R2 in all three fertility stages was 0.87, and the root mean square error (RMSE) was 0.70. Compared with the original YOLOv5s algorithm, the spikelet detection counting effect of YOLOv5s-T was not reduced. Meanwhile, the model size was reduced by 36.8% (only 9.1 M), the GPU memory usage during the training process was reduced by 0.82 GB, the inference time was reduced by 2.3 ms, the processing time was reduced by 10 ms, and the calculation amount was also reduced. The proposed YOLOv5s-T algorithm significantly reduces the model size and hardware resource requirements while guaranteeing high detection and counting accuracy, which indicates the potential for wheat spikelet counting in highly responsive wheat yield estimation.
Abstract from the paper’s OpenAlex record; open copy licensed cc-by. Read at the source ↗
The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.
Claims 11
Claim · model-written summary
Predicted and true flowering stage spikelet counts had R2 of 0.97.
- Taxon
- wheat
Triticum aestivume0 - Organ / stage
- flowering stagee0
- Outcome
- coefficient of determination (R2)e0
- Value
- 0.97e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“coefficient of determination (R2) between the predicted and true values of wheat spikelets was 0.97 for the flowering stage”
Claim · model-written summary
Predicted and true filling stage spikelet counts had R2 of 0.85.
Not reportedTaxon
Verbatim
e0“0.85 for the grain filling stage”
Claim · model-written summary
Predicted and true mature stage spikelet counts had R2 of 0.78.
Not reportedTaxon
Verbatim
e0“0.78 for the mature stage”
Claim · model-written summary
Across all three stages, R2 was 0.87.
Not reportedTaxon
Verbatim
e0“The R2 in all three fertility stages was 0.87”
Claim · model-written summary
Across all stages, RMSE was 0.70.
Not reportedTaxon
Verbatim
e0“the root mean square error (RMSE) was 0.70”
Claim · model-written summary
YOLOv5s-T spikelet detection and counting performance did not decline versus original YOLOv5s.
- Outcome
- spikelet detection counting effecte0
- Direction
- not reducede0
- Comparator
- original YOLOv5s algorithme0
Not reportedTaxon
Verbatim
e0“Compared with the original YOLOv5s algorithm, the spikelet detection counting effect of YOLOv5s-T was not reduced”
Claim · model-written summary
Model size decreased 36.8% to 9.1 M.
Not reportedTaxon
Verbatim
e0“model size was reduced by 36.8% (only 9.1 M)”
Claim · model-written summary
Training GPU memory use decreased by 0.82 GB.
Not reportedTaxon
Verbatim
e0“GPU memory usage during the training process was reduced by 0.82 GB”
Claim · model-written summary
Inference time decreased by 2.3 ms.
Not reportedTaxon
Verbatim
e0“inference time was reduced by 2.3 ms”
Claim · model-written summary
Processing time decreased by 10 ms.
Not reportedTaxon
Verbatim
e0“processing time was reduced by 10 ms”
Claim · model-written summary
Calculation amount also decreased.
Not reportedTaxonComparator
Verbatim
e0“calculation amount was also reduced”