Introducing
Your new presentation assistant.
Refine, enhance, and tailor your content, source relevant images, and edit visuals quicker than ever before.
Trending searches
10
Date:
18/01/2020
code4cause
Ministry Category
ISRO - Software
Introduction
Problem Code
NM378
Problem Statement:
Develop a mobile application that can identify crop using only field photo of a crop. The team must target at-least 10 different crops for demonstration. The application will allow the user to take photos and automatically identify the crop. The photo and crop information along with Geo-location information should be stored in an internal database which can be exported/emailed.
Introduction
College Code
126
Team Leader
Nishit Mistry
Why do we need this system?
Need
1. Smart phones today can take the high quality pictures, thus our system uses these pictures to compare the attributes of different crops and to get the detailed information about them.
2. Deep Learning Algorithms can be used which provide more convenient and accurate data with faster and efficient results.
What are the existing systems?
Existing System
1. There are many existing systems which have been implemented in agriculture and they prove to be very useful to farmers.
2. But, these systems are based upon IoT technology to gather the knowledge about the crops.
3. Thus this is a very long process, and not scalable as it requires high maintenance.
4. So there is a need to develop a solution which is scalable and should reach to a majority of people.
Solution/Idea
Solution/Idea
1. The system is based on object detection which
is achieved through Deep learning algorithm- Convolutional Neural Networks (CNN) algorithm.
2. The system is trained for identification of 12 different crops using anaconda environment and with help of TensorFlow object detection API.
3. The system is then tested to avoid any misjudgment.
4. When the Scanner scans the crop, the crop is identified into one of the 12 classes.
5. The details of the identified crop is displayed on the screen by retrieving the information stored in the database and location coordinates are displayed on the map.
6. This data can be exported/emailed which can be used by the authority to plan some development scheme for farmer in that particular area.
How this system will be better?
Efficiency
1. ML & DL provide a better solution as compared to the conventional method or manual method.
2. This system can help the scholar in the field of agriculture to get more information about the crops.
3.It will also help someone who is novice and has an interest in agricultural science and wants to know crops, their patterns, etc.
4.This system also helps the governing authority to analyze a region geographically in terms of crops cultivated with the data which it has got through the mobile application.
Analysis
Dependencies
Analysis
Cost Analysis
Cost Analysis
1. Minimum requirements for the system to work properly:
-Windows 10 operating system
-Intel i5 8th generation processor
- 8 GB RAM
- 1 TB Hard Disk
-NVIDIA Geforce GTX 1050ti 4 GB
A system with the above configuration may cost around 60,000 INR.
2. App registration fees on Google Playstore is $25 i.e. 1,772.44 INR.
3. Firebase pricing starts at $24.99 per month i.e.
1,771.73 INR.
4
Fig1. Flow Diagram
Technology stack
Technology Stack
1. Android Studio version: 3.5.3
2. Programming languages: Java, Python & XML
3. Toolset: Android Studio and Android Developer Tools
4. TensorFlow version: 1.15.0
5. TensorFlow Lite
6. Anaconda environment with python version: 3.6.9
7. Google Maps and Firebase API
Use case
Fig2. Use Case Diagram
Conclusion
Deep learning and Machine Learning has always been the part of the high computing machine, bringing it to the mobile devices increases the usefulness of this powerful technology. This system is best example of this implementation. Agriculture has employed 50% of the Indian work force and contributed 17–18% to country's GDP. This number can further be increased if the interest of the youth is harnessed towards the farming. In the age of technology the tools like Cropifier will help the enthusiast to explore this field.
Conclusion
Thank You
What's Next?
Future Scope
1. With successful working of Cropifier, our app can be further modified to identify not only multiple different crops but also varieties of plants, flowers and trees.
2. App users could access the information about variety of different plants, their medicinal properties, type of soil required, suitable weather conditions for their growth, etc.
3. This system can be further refined with support of many regional languages.
4. More efficient Deep Learning Algorithms can be used to train the model with even less time and more accuracy.
5. Cropifier could also be extended in such a way that it could also work in iOS devices.
4
Thank You!