Evidence brief: camera observation versus diagnosis
Which claims can a plant image, model score and laboratory result each support?
Scope of this review
Visible-light plant images, interpretation of model evaluation and the diagnostic role of case context. No validation of a Hamfy model or performance claim for cannabis.
How we assessed the evidence
We compared the question addressed by each source, the material examined and the limits of the conclusion. The engineering interpretation is Hamfy’s; the underlying experiments belong to the cited authors.
What the evidence supports
Transfer is an empirical question
The published PlantVillage experiment illustrates that performance changes when image conditions change.
Different evidence answers different questions
A photograph documents appearance; diagnostic methods add information about a defined sample and target.
What this review does not establish
- The 2016 model is not representative of every current system.
- The cited image experiment does not validate cannabis disease recognition.
- No image collection, laboratory comparison or accuracy measurement was performed by Hamfy for this article.
Hamfy’s proposed next step
Keep observations, measurements and confirmed findings distinct in the same record. A monitoring interface should make the origin and uncertainty of each claim visible.
Evidence to collect for a real project
- The intended decision and the evidence needed to support it.
- Traceable image, sensor and laboratory records from the relevant setting.
- Independent review of diagnostic claims and performance on unseen cases.
Source notes
- Research source · 2016Using Deep Learning for Image-Based Plant Disease DetectionMohanty, Hughes & Salathé · Frontiers in Plant Science
Primary image-classification experiment; controlled and external evaluation are reported separately.
- Research source · 2024Field Notes talks crop disease diagnosis resourcesUniversity of Minnesota Extension
University plant-diagnostic guidance on symptoms, context and laboratory investigation.