Crop Health is Laisha's photo-based crop diagnosis: photograph what is wrong, describe what you have already noticed, and get back an organised reading of what is actually visible, what it could be, and how confident any of that is.
It accepts several photographs of the same problem and reads each one in turn, because one leaf rarely tells the story. A pattern that looks like disease on a single leaf often turns out to be the shape of the affected area in the field — worse at the edges, worse in the low corner, worse on one variety — and that is visible only across several pictures.
It deliberately separates three things that most advice runs together: what can be seen in the photograph, what might be causing it, and how uncertain that is. A photograph alone is never treated as confirming a disease or prescribing a treatment. Anything that would cost money to act on is flagged as needing a person who can stand in the field.
It asks for the crop, the growth stage, where the pattern shows up in the field, recent weather, irrigation and what field operations were done lately. Those are the questions a good agronomist asks before looking, and they change the answer — the same yellowing means different things at tillering and at grain fill, and after a spray rather than after a flood.
It then builds a crop-health plan rather than returning a name: what to check next, what to rule out first, and what to do if the cheapest explanation turns out to be right. A diagnosis nobody can act on is a word, not help.
It is one of the paid tools, and it is honest about why: reading photographs costs real money per use. The ten free days are enough to test it on a problem you already know the answer to, which is the right way to decide whether to trust it on one you do not.
How to use Crop Health
What it is. A crop health check from a photograph. Send one whole-plant photo and one close-up of the affected part, add whatever you know about the crop and recent conditions, and it comes back with what the evidence shows, the possible cause groups, what to do in the next 24 hours, a seven-day watch plan, and a brief you can hand to an agronomist.
When to open it. The day you first notice something, not the week after. Almost every crop problem is cheaper and more treatable early, and the photographs are better before the damage spreads.
Have ready.
A clear photo of the affected plant or leaf, in focus.
Ideally a second, wider photo of the whole plant — the two together give a far better answer than either alone.
The crop name.
Optionally: crop age or stage, how much of the field is affected, how serious it looks, and what the weather and the spray and fertiliser history have been.
Step by step.
Take the main photo of the affected leaf, stem or fruit, in focus, in daylight.
Add a close-up. The screen asks for both because a symptom's pattern on one leaf and its distribution across the plant are two different pieces of evidence.
Enter the crop name — wheat, cotton, tomato.
Fill in the optional fields. Each one narrows the answer: 45 days and flowering rules out a great deal that 'wheat' alone does not.
Say how widespread it is — a few plants, patches, or the whole field — and how concerned you are.
Press 'Analyze crop health' and read the report. Download the PDF if you are taking it to somebody.
Use the 'Message for your agronomist' section. It is written to be forwarded.
What it will not do. It reads photographs and what you tell it. It cannot test tissue, cannot see roots, cannot culture a pathogen, and it does not prescribe a chemical or a dose. It gives cause groups and safe immediate steps, and it tells you when to stop using it and call somebody.
Every setting in Crop Health, explained
Main photo
The primary evidence. Everything else is context around it. How to choose: Daylight, in focus, the affected part filling the frame. Photograph the boundary between healthy and affected tissue if you can — that edge carries most of the diagnostic information. If it is wrong: A blurred or dark photo produces a vague answer, and the report will say the evidence was weak rather than guess confidently.
Close-up
A second, tighter photo. How to choose: Get close enough to see whether the spots have rings, whether there is anything growing on the underside, whether the edges are sharp or diffuse. If it is wrong: Omitting it is the single biggest reduction in answer quality, and the screen says so: best result is one whole-plant photo plus one close-up.
Crop name
Which crop it is. How to choose: The common name is enough. Add the variety in the description if you know it and it is locally distinctive. If it is wrong: Nearly every diagnosis is crop-specific; without this the report is reduced to describing what it can see.
Crop age or stage
How old the crop is, or what stage it is at — '45 days, flowering'. How to choose: Either is fine. Stage is more useful than age where you know it. If it is wrong: Many problems only occur at particular stages, so this both adds candidates and removes them.
How widespread (a few plants, patches, whole field)
The distribution of the symptom across the field. How to choose: Walk and look before answering. This is diagnostic in itself. If it is wrong: Distribution separates whole categories of cause. Scattered single plants suggests something different from sharp patches, which suggests something different again from a uniform whole-field symptom — which is very often not a disease at all but nutrition, water or a spray.
Concern level (monitor, act soon, urgent)
Your own judgement of how serious it is. How to choose: Honestly. It shapes how the 24-hour advice is pitched. If it is wrong: Understating it to avoid alarm gets you advice for a problem you do not have.
Recent conditions
Rain, heat, irrigation, fertiliser or spray in the recent past. How to choose: Especially any spray, and especially a herbicide — on your field or the one next to it. Drift damage is common and looks like disease to anyone who does not know the spray went on. If it is wrong: This field probably changes more diagnoses than any other optional one. A symptom that appeared three days after a spray is a very different case from one that appeared after a week of rain.
Describe it instead
A text description, where there is no photo. How to choose: Be concrete — 'yellow lower leaves, brown edges' beats 'looks unhealthy'. If it is wrong: A description-only answer is necessarily broader, and the report will be honest about that rather than pretending to have seen the plant.
A worked example
Yellowing on the lower leaves of a 45-day tomato crop, in sharp patches across part of the field, a week after heavy rain.
What goes in:
Main photo plus close-up
Crop: tomato
Stage: 45 days, flowering
Distribution: patches
Recent: heavy rain, no spray
What comes out:
What the evidence shows: the symptom pattern visible in the photographs, stated as observation rather than conclusion.
Possible cause groups, ordered — with patches after heavy rain putting waterlogging and root-zone problems alongside foliar disease rather than behind it.
What to do in the next 24 hours: safe steps that do not commit you to a diagnosis.
A seven-day watch plan: what to look for, and what would change the answer.
Get expert help if: the specific signs that mean stop waiting.
Notice what it does not do. It does not name one disease and tell you to buy a chemical. The 24-hour steps are deliberately the ones that are safe whichever of the cause groups turns out to be right.
How to read what it gives you
The report separates what it observed from what it inferred, and it says which is which. That distinction is the difference between evidence and a guess.
Cause groups, not a single named disease. A photograph genuinely cannot distinguish between several causes that look alike, and a confident single answer would be the dishonest way to present the same information.
The 24-hour advice is chosen to be safe across the candidate causes. Anything that commits you — a specific chemical, a specific dose — is left to somebody who can stand in the field.
'Get expert help if' is the most important section on the page. Read it first if the crop matters.
What usually goes wrong
One blurred photo and no context. The tool is only as good as its evidence and it will tell you the evidence was thin.
Photographing only the worst plant. The spread across the field is diagnostic; photograph the edge of the affected area too.
Leaving out a recent spray. Herbicide drift is common and mimics disease.
Acting on a cause group as though it were a diagnosis, and buying a chemical for it. Use the agronomist brief — that is what it is for.
Waiting a week to check, by which time the early, cheap, treatable stage is gone.
Questions people ask about Crop Health
What are the yellow spots on my crop's leaves?
That depends on the pattern, the crop, the stage and what has happened recently — which is why the tool asks for all four rather than answering from the colour alone. Photograph the boundary between healthy and affected tissue, add a close-up, and say what sprays and weather preceded it; those four things separate most of the common causes.
How should I photograph a crop problem?
In daylight, in focus, with the affected part filling the frame, plus a second wider photo of the whole plant. Include the edge where healthy tissue meets affected tissue — most of the diagnostic information is in that boundary. Two photos are substantially better than one.
Will it tell me which pesticide to spray?
No. It gives cause groups and steps that are safe whichever of them is right, then tells you what evidence would settle it. A chemical recommendation needs a confirmed diagnosis, your local registrations, the crop's stage and a pre-harvest interval, and a photograph supplies none of those.
Can I use it without a photo?
Yes — describe what you see instead. The answer will be broader, and the report will say so rather than pretending to have seen the plant.
What does the seven-day watch plan do?
It tells you what to look at over the coming week and what would change the answer — whether the patches spread, whether new leaves are affected, whether it stops at a field boundary. That last one, in particular, distinguishes a spray problem from a disease.
Why does it give several possible causes instead of one answer?
Because several real causes look alike in a photograph, and narrowing to one would mean choosing arbitrarily and sounding certain about it. Cause groups plus the evidence that would separate them is more useful to act on, and it is what an agronomist would say standing in the same field.
Want to use Crop Health for real, right now? Open Crop Health in the Laisha app → Ten days free, no card. After that 10 US dollars a month, one price worldwide.