Plantix Vision

Leverage the power of Plantix Vision for your agtech application

Receive instant diagnosis from our AI image recognition technology!

Integrate Plantix Vision into your mobile offering!

Leverage the power of Plantix Vision for your agtech application
Receive instant diagnosis from our AI image recognition technology!
Integrate Plantix Vision into your mobile offering!

Plantix Vision Disease Recognition Offers...

{{ translate(item, 'business-offers-argument-', '-title')}

350+ Plant Diseases

Automated diagnostics for over 350 globally identified plant diseases, pests and nutrient deficiencies on more than 65 agricultural crops

A Clear Advantage

The neural networks of Plantix Vision are trained on the biggest plant and disease image database worldwide with more than 20k new images per day.

Real-Time Updates

The Plantix Vision database is constantly improving its diversity and quality. Benefit from the continuous advancements too.

Deliver your customers the best plant disease recognition service out there.

Provide them with detailed information and help their business to thrive.

Show Documentation

Features

Disease Recognition

By specializing in detecting plant pathogens, Plantix Vision makes cutting-edge image recognition technology available for agrarian purposes and provides understandable results with internationalization. Profit from an outstanding and field proven accuracy ensured by three independent working neural networks.

EPPO Code
English Disease Name
Scientific Name
Similarity As Percentage
JSON Format

Disease Net

The disease net is trained on hundreds of thousands of images from more than 40 countries and is the first model on the market that can detect more than 350 plant damages.

Crop Net

The crop net is the model with the highest accuracy for crop detection even with complicated light settings and in difficult environment. The crop and disease model work as a team, so anytime a disease is recognized, the crop net will identify the plant that is affected.

Object Net

The object net is especially meaningful when people trying the app for the first time and simply snap images from arbitrary objects in their direct surrounding. In this case, the object net tells you what exactly the user snapped. (coffee mugs, television, their feet)

Convince yourself!

150 free API calls!

See For Yourself...

Plantix Vision recognizes 350+ plant pathogens on 66 major agricultural crops. 5 years of field use have proven its accuracy, versatility and reliability.

Try the leading image recognition API for plant pathogens now! Packages starting at €15000 per year!

Embrace New Possibilities

Support Your Customers

By equipping your agricultural extension workers, employees, and customers with Plantix Vision, you can easily tweak their workflows. So your personnel can spend more time on practical advisory and less on disease identification.

Boost Your App With AI

By integrating our API into your agricultural service application, you can upgrade your weather, market or disease information service into a fully comprehensive advisory tool and provide your user base new valuable services.

Get Statistical Insights

Whether for the purpose of quality assessment or for monitoring large studies, Plantix Vision enables you to raise a new level of data acquisition to gain statistical insights for increasing efficiency in your agricultural business.

Get an easy-to-use API response.

Example 1: Diseased Plant

When Plantix Vision recognizes a single disease, the image_analysis array will contain only one object that provides the information about the disease. The plant_net will provide the recognized crop. And a Boolean value for the recognition will be assigned the value true.

JSON Response

{
    "big_enough": true,
    "code": 200,
    "image_analysis": [
        {
            "eppo": "PYRIOR",
            "name": "Blast of Rice",
            "peat_id": 100058,
            "rank": 1,
            "scientific_name": "Magnaporthe oryzae",
            "similarity": 90
        },
        {
            "eppo": "RHIZSO",
            "name": "Rice Sheath Blight",
            "peat_id": 100080,
            "rank": 2,
            "scientific_name": "Rhizoctonia solani",
            "similarity": 7
        }
    ],
    "objects_net": [
        {
            "name": "stinkhorn",
            "rank": 1,
            "similarity": 89
        },
        {
            "name": "obelisk",
            "rank": 2,
            "similarity": 2
        }
    ],
    "pic_id": "2ea1cf46-4c23-4654-b503-c8c9d6b0fc75",
    "plant_group_net": [
        {
            "name": "RICE",
            "rank": 1,
            "similarity": 100
        }
    ],
    "plant_net": [
        {
            "name": "RICE",
            "rank": 1,
            "similarity": 99
        }
    ],
    "recognized": "True",
    "recognized_bool": true
}

Example 2: Crop Group Recognized

Some crops look very similar in their early stages, millet, sorghum and maize would be a classic example. To avoid false positives, our image recognition nets will always return the crop group with the same probability. In this way, it is easy for our customers to ask their users exactly which crop is shown in the picture and then get valuable feedback directly about possible pathogens.

JSON Response

{
    "big_enough": true,
    "code": 200,
    "image_analysis": [
        {
            "eppo": "PYRIOR",
            "name": "Pyricularia Leaf Spot",
            "peat_id": 100120,
            "rank": 1,
            "scientific_name": "Magnaporthe oryzae",
            "similarity": 82
        },
        {
            "eppo": "COLLGR",
            "name": "Anthracnose Leaf Blight",
            "peat_id": 100084,
            "rank": 2,
            "scientific_name": "Colletotrichum graminicola",
            "similarity": 6
        }
    ],
    "objects_net": [
        {
            "name": "ear",
            "rank": 1,
            "similarity": 86
        }
    ],
    "pic_id": "6bcc6cc4-9249-4541-89f6-7221639bf046",
    "plant_net": [
        {
            "name": "MAIZE",
            "rank": 1,
            "similarity": 97
        },
        {
            "name": "MILLET",
            "rank": 1,
            "similarity": 97
        },
        {
            "name": "SORGHUM",
            "rank": 1,
            "similarity": 97
        }
    ],
     "recognized": "False",
    "recognized_bool": false
}

Example 3: Crop recognized

In some cases the API can’t be sure about the possible pathogen that is affecting the plant. The plant could look healthy or the damage is not directly visible on the submitted picture. In another case, the image is not taken accurately. Plantix Vision will try to avoid any false positives and will only shoot for a recognized crop, but not the exact pathogen. In this case it would show a ranked result list to select the pathogen, but return a -recognized false- as Boolean value.

JSON Response

{
    "big_enough": true,
    "code": 200,
    "image_analysis": [
        {
            "eppo": "",
            "name": "Healthy",
            "peat_id": 999999,
            "rank": 1,
            "scientific_name": "Healthy",
            "similarity": 72
        }
    ],
    "objects_net": [
        {
            "name": "lemon",
            "rank": 1,
            "similarity": 78
        },
        {
            "name": "orange",
            "rank": 2,
            "similarity": 5
        },
        {
            "name": "hip",
            "rank": 3,
            "similarity": 2
        }
    ],
    "pic_id": "9dd02933-c081-4718-aeba-79f309df0468",
    "plant_net": [
        {
            "name": "GRAPE",
            "rank": 1,
            "similarity": 99
        }
    ],
    "recognized": "False",
    "recognized_bool": False
}

Example 4: Object Recognized

Often users do not really send a picture of a plant, let alone a detectable pathogen. In these cases, Plantix Vision always delivers the result of our object network. Instead of -no plant- or -unidentifiable-, our API returns an intelligent result to the user. More than 1000 objects can be detected with high accuracy. In this way, Plantix Vision always offers the ability to provide a reasonable result, even if the user is just trying to challenge the IR.

JSON Response

{
    "big_enough": true,
    "code": 200,
    "image_analysis": [
        {
            "eppo": "UNKNOWN",
            "name": "Object",
            "peat_id": 0,
            "rank": 1,
            "scientific_name": "UNKNOWN",
            "similarity": 99
        }
    ],
    "objects_net": [
        {
            "name": "lighter",
            "rank": 1,
            "similarity": 85
        },
        {
            "name": "syringe",
            "rank": 2,
            "similarity": 4
        }
    ],
    "pic_id": "bfbc8e90-3e46-4ee2-996d-ed77ae280ea3",
    "plant_group_net": [
        {
            "name": "EGGPLANT",
            "rank": 1,
            "similarity": 100
        }
    ],
    "plant_net": [
        {
            "name": "EGGPLANT",
            "rank": 1,
            "similarity": 22
        },
        {
            "name": "BANANA",
            "rank": 2,
            "similarity": 14
        },
        {
            "name": "RICE",
            "rank": 3,
            "similarity": 11
        }
    ],
    "recognized": "False",
    "recognized_bool": false
}