Abstract
• Smartphone photos enable scalable coffee yield estimation in real field conditions. • Branch segmentation and occlusion correction improve yield estimation accuracy. • Vision pipeline performs well across smallholder systems and data capture modalities. • Data capture modalities impact both yield accuracy and ease of image collection. Accurate coffee yield estimation is critical for crop management, labor and financial planning, and value-chain transparency, including compliance with the EU Deforestation Regulation. However, manual cherry counting remains labor-intensive, error-prone, and unreliable as worker fatigue sets in, highlighting the need for automated, scalable alternatives. This study introduces a novel deep-learning framework for automated coffee cherry counting using images captured with low- to mid-range smartphones across diverse smallholder farming contexts. The pipeline combines automated branch segmentation, cherry detection, and a regression-based correction module to account for occluded cherries, accommodating different data-capture modalities. We evaluated the framework on 7,025 annotated images from Colombia, Peru, Honduras, and Uganda, covering both Coffea arabica and Coffea canephora (Robusta) coffee species. Under optimal image-capture conditions (i.e., full background isolation), the model achieved high accuracy, reaching an R 2 of up to 0.96 and reducing the Mean Absolute Percentage Error (MAPE) to as low as 10% at the plot level, outperforming state-of-the-art methods. By reducing manual effort and addressing real-world constraints in smallholder settings, this approach offers a strong foundation for scalable coffee yield estimation. Future research should prioritize human-centered design validation and detailed cost-benefit analyses to support widespread adoption and long-term sustainability.
Abstract from the paper’s OpenAlex record; open copy licensed cc-by-nc-nd. 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 4
Claim · model-written summary
Branch segmentation and occlusion correction improved coffee yield estimation accuracy.
CoffeaNot reportedComparator
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Branch segmentation and occlusion correction improve yield estimation accuracy”
t0“coffee yield estimation”
Claim · model-written summary
Data capture modalities affected coffee yield accuracy and ease of collection.
- Outcome
- yield accuracy and ease of image collectione0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Data capture modalities impact both yield accuracy and ease of image collection”
t0“coffee yield estimation”
Claim · model-written summary
At plot level under full background isolation, the model reached R 2 up to 0.96.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Under optimal image-capture conditions (i.e., full background isolation), the model achieved high accuracy, reaching an R 2 of up to 0.96”
t0“coffee yield estimation”
Claim · model-written summary
At plot level under full background isolation, MAPE was as low as 10%.
- Outcome
- Mean Absolute Percentage Error (MAPE)e0
- Direction
- reducinge0
- Value
- 10%e0
- Comparator
- state-of-the-art methodse0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“reducing the Mean Absolute Percentage Error (MAPE) to as low as 10% at the plot level, outperforming state-of-the-art methods”
t0“coffee yield estimation”