Plant Health Assessment

Submit photos of a sick plant to our backend and get the JSON result with plant health status and specific problems.

See the Plant.id Health Assessment product page.

Request

Endpoint:

https://api.plant.id/v2/health_assessment

Request method:

POST

Required parameters:

images - a list of one or more images of the plant you want to diagnose (images should be encoded as strings in Base64)

Optional parameters:

latitude - geographic coordinate (float) - increases the identification accuracy
longitude - geographic coordinate (float) - increases the identification accuracy
modifiers - list of strings that influence the identification process and results (speed, accuracy, response content), see the list of available modifiers
disease_details - list of strings that specify what additional information about the plant disease (e.g. description, treatment) will be included in the response, see the list of available disease details
language - language code (ISO 639-1, default "en") that specifies the language of the response (supported languages: English - "en", German - "de", Czech - "cs", Spanish - "es", French - "fr", Italian - "it", Dutch - "nl", Polish - "pl", Swedish - "sv", Chinese - "zh", Danish - "da", Turkish - "tr", Hindu - "hi", Arabic - "ar", Portuguese - "pt", Korean - "ko")
languages - list of up to 3 language codes (ISO 639-1) to specify languages of the response (do not combine with the language parameter); if you use this parameter, the corresponding fields in the response will be converted to dictionaries with the language code as the key.
custom_id - unique identifier you can set for your purpose (int) - can be used to handle network connection errors
datetime - timestamp in seconds (int)
identification_timeout - timeout limit in seconds (int, default 20 seconds, max. 20 seconds); if the identification takes longer than the given number of seconds, the identification info is returned without a suggestion
prune_diseases - boolean (optional) - if true: list of suggested diseases is limited to more specific diseases with general diseases such as "Abiotic" or "Water-related issue" removed.

See the request examples in different programming languages in our repository.

Response

The result includes plant health status and a list of suggested diseases in JSON format. There are 9 root categories: Abiotic, Animalia, Bacteria, Canker, Chromista, Dead plant, Fungi, Senescence, and Viruses. Each suggestion includes:

entity_id - a unique identifier for each disease that will not change over time
name - the name of the disease in our database (Latin or English), which may change over time
probability - level of certainty that the suggested disease is present in the image
redundant - a boolean value indicating that the class can be discarded when the difference between the predicted probability of parent (higher in the hierarchy) and child class (lower in the hierarchy) is insignificant; specifically, this feature in the response activates when the probability of the child class is at least 80% of the probability of the parent class (or less than two percentages in absolute numbers that aims to assess cases with a probability of less than 10%); for example, when the probability of a parent class (e.g. abiotic) is 95% and the probability of a child class (e.g. water-related issue) is at least 76%, then the attribute redundant:true will appear in the response
similar_images - a list of representative images of the identified disease selected by the model to resemble the input image; the list contains dictionaries with image data (id, similarity, url, url_small); similar images are included in the result only if you add the value similar_image in the modifiers list in the request
disease_details - info you requested (if available) about suggested diseases, see Disease details documentation page

The information about health status contains a boolean attribute is_healthy, which indicates whether the plant is healthy or sick, and is_healthy_probability, which estimates the probability that the plant is sick.

Response example

See the JSON response example.

Asynchronous response retrieval

You can later retrieve the identification results using the identification ID or CUSTOM_ID. Make sure that identification with the given ID (or your CUSTOM_ID) has already been processed.

URIs: