Most frequent questions and answers
- if possible, include multiple photos of a single plant to the plant identification request
- if possible, include GPS coordinates
- There is parameters list (this is a request payload an attribute in in /identify-plant endpoint API call). By putting some keywors in there, it’s possible to tread off some plant identification speed to reach a better results. It’s no alloed by default, so please contact us, we’ll help.
- The default model is very general. Bu we are able to train a specific model focused just to your use case — for example mediterranean crops, Canadian weeds or European trees. We can even your your own datasets to go beyond the identification task like estimating the plant’s health, or improve the model with satelite images.
Our machine learning model is based on the photos we gathered through FlowerChecker database. Most training data contains the photos of both wild, garden and indoor plants. We can currently identify more thank 10 000 plant species (trees, shubs, succulents and some most common musrhooms and mosses included).
The API will provide the best results with the plants from the green countries.
The base pricing is 5 eurocents per identification. For large orders it price might be lower, it depends on the balancing the speed, and accuracy. Please contant us on firstname.lastname@example.org we can discuss your needs.
The name of the plant is not the only piece of information you’ll get from the API. We’ll also provide you with representative image of the suggested taxon (specie). You are free to use this taxon in your app.
But hese representative images are not static, but adaptive in order to achieve better user’s response. Let me illustrate
These are three different identifications of three different plants of Dandelion specie. What makes them different is vegetation stage of the plant (puff ball, leaves only, full bloom). As you can see the Plant.id reacts on the the user’s picture and provides a different representative images of the Dandelion so it’s actually similar to what user sees.
This is a great tool to improve user’s satifcation with thes service.