Rashi Srivastava, IIIrd Semester - [email protected]
Ankit Jailwal, IIIrd Semester - [email protected]
Shivam Sawarn, IIIrd Semester - [email protected]
Abhisehk Kumar, IIIrd Semester - [email protected]
Vivek Kumar, IIIrd Semester - [email protected]
Selecting Crops for farmers is a hectic job as they have to consider many different factors. Hence, developing an app which will help farmers in making their jobs easy.
The aim is to design an AI based crop recommendation application which can recommend different crops to the farmers by analysing all the relevant factors for a particular land area such as rainfall, temperature, season, ground water available, soil type and location. The application is based on Deep Learning and Machine Learning algorithms to detect the soil type given an image and recommend the best fit crop along with further suggestions for that very crop such as expected revenue generated per hectare, demand of the crop, required fertilizers, cost of cultivation per hectare, quantity of seeds per hectare, duration of cultivation and the crops which can be used for mixed cropping with the primary crop.
Multiple datasets have been used for training and obtaining relevant output purposes. All the datasets used are custom datsets; constructed and structured according to the algorithm requirements and proposed test cases. Following is the list and types of the datasets used-
1.) Soils.zip(Soil Image Dataset) - Contains around 600-700 images of different soil types which are used for agriculture and generally found in Indian-Subcontinent.
2.) Cat_crop.csv - The mentioned CSV file contains data related to various parameters which have been taken into account while training the Crop recommendation System Machine Learning Model.
3.) Prediction.csv, Prediction.json - This dataset contains additional information about the crops used in training of the model and are fetched in real time after crop prediction to give the user a better idea about the requirements of crops and their needs.
-
The app starts after an easy user login using mobile number but for the prototype the mobile number feature hasn't been implemented completely and the app is currently working on a dummy user name and phone number.
-
After login the user is taken to the main menu page where, the user can navigate through different features of MAATI APP.
-
For Crop prediction, the user is required to click the image of the land on which the farming has to be done and send it to the server for processing and to predict soil type using Deep learning model.
-
The predicted soil type will be used as a parameter for crop recommendation.
-
User's current location and weather conditions will also be sent to the server for crop prediction and altogether all the parameters will be used to predict the crop which is suitable for cultivation.
-
After crop prediction the relevant details related to the crops will be fetched and shown as the output to the users.
-
Users can also access other features of the app such as: Kisan Call Centre, Maati News for agriculture news and happenings, Maati analysis for weather updates, History of previous recommendations etc.
Enter your username and phone no. to register.
As you enter your details ou will enter to the page which contains three sections buy, sell and rent.
In buy section you may buy all the things that a farmer needed.
On clicking on buy section you land at a page which contains six section name as maati shop , maati news , maati analysis , kisan call center , history , contact us.
All the things needed by a farmer may be found here such as seeds , fertilizers , tools and pesticides.
You may select your item.On clicking on buy now you will be directed to razorpay payment gateway .Further you may buy these things oncash, paylater and easy emi.
Keep you update with the things related to farmers.It get updated daily.
Farmer can take the picture of soil through their camera and maati app will analyze the whole thing for it such as suitable weather condition for soil, suitable crop etc.
This will directly direct them to their call section to call with kisan seva kendra contact no.
History of what transcations made, purchased, sell and rent will kept here .
Enter the name of item you want to sell with description, quantity and amount . Capture its snap and sell now button will post it to the needed ones.
Enter the name of item you want to rent with description, quantity and amount . Capture its snap and rent now button will post it to the needed ones.
It will direct you to a sidebar contains all features such as maati shop, news etc.
It will allow you to choose the diffrent languages.
For convenience we are providing night mode.
We giving our basic information at that page.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent