Smart Agriculture with Deep Learning

Project Title

Smart Agriculture with Deep Learning

Project Area of Specialization Artificial IntelligenceProject Summary

Agriculture is considered the spine of Pakistan's economic system, which relies on its important vegetation. Pakistan is among the biggest nations which belong to the agriculture gadget and develop vegetation in a huge variety like cotton, sugarcane, rice, oranges, mangoes, potatoes, onions, etc. Having diseases in plants & crops are quite natural. If proper care is not taken in this area then it causes serious effects on plants due to which respective product quality, quantity or productivity is affected. Problem is the detection of disease. Plant diseases are in general classified in three most important training inclusive of viral, bacterial and fungus. Farmers are not well educated and don’t have the knowledge about diseases that damage the crops. Detection of plant disease through some automatic technique is beneficial as it reduces a large work of monitoring in big farms of crops and at a very early stage. In this work, an application will be developed using image processing and deep learning techniques that will not only perform species identification from leaf image but also detect either given image belongs to normal or abnormal crop and in case of a diseased leaf, it suggests a remedy.

Project Objectives

The main objective of this project is to develop an application for leaves disease detection in order to increase crop’s/plant’s productivity and quality which directly benefits human health. This software will be able to

•    Identifies the species of the crop.
•    Detect the diseased leave and type of disease.
•    Suggest the remedy of the disease.

Project Implementation Method

To achieve the desired objectives of the project the following will be the major steps:

Step 1: Data Gathering and Labeling

Initially, data of normal and abnormal leaves belonging to numerous crops are collected. After that dataset will be labeled with spices type and type of diseases. 

Step 2: Convolutional neural network (CNN) Training

Convolutional neural network (CNN) is trained for leaf species detection and to decide either given leaf is normal or abnormal. Moreover, in the case of abnormality, it tells the type of abnormality. For network training, the Python programming language is used along with tensor-flow API. The dataset contains leaves of different species of plants for network training and testing. As CNN training is a computationally intensive task, so in order to train it, Graphical Processing Unit (GPU) will be used  

Step 3: Application Development

An Android application will be built that takes an image of leaves as input and feed it to the trained CNN model that will identify crop species, detects either the leave is infected or not and finally provide/pronounce the remedy of disease.

Benefits of the Project

The following will be the advantages of our project :

1. Increase the productivity and quality of crops/plant.
2. Save the user time a user does not need to visit an expert to get the required solution to the problem.
3. Easy to use as the user get the solution in just a click
4. Help to improve the economy of the country. 
5. Help to uplift the standard of life.

Technical Details of Final Deliverable

The final deliverable will be the android application that recognizes the crop species, detect either crop is normal or abnormal, in case of abnormality tell the type of abnormality and suggest the remedy.

Final Deliverable of the Project Software SystemType of Industry IT , Medical , Agriculture Technologies Artificial Intelligence(AI)Sustainable Development Goals Zero Hunger, Good Health and Well-Being for People, Decent Work and Economic Growth, Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	70000
GPU	Equipment	1	40000	40000
Android Device	Equipment	1	30000	30000