Automatic Plant Leaf Disease Detection and Auto-Medicine Using IoT Technology

Abstract

Leaf diseases in plants cause substantial production and economic losses, as well as a decrease in both the quality and quantity of the crops.  It is easier to identify leaf diseases early in leaf health and can promote disease control through proper management strategies. This chapter provides a method for the early detection of leaf diseases in plants based on some important features taken from their images. Beaglebone Black (BBB) is interfaced with a digital camera used to take images of plant leaves to detect diseases in leaves. The image of the leaves is captured in the proposed framework and contrasted with images in the leaves database which are pre-stored. After image comparisons, if the plant leaves are found to be infected, this device automatically supplies medicine by a sprinkler to the area of the plant leaves. Leaf diseases are identified by initial or final phase tests. The responsible administrator  may allow the medicine by sending a message back to GSM (Global System for Mobile Communication) if necessary. The diseases are at the final stage,  the system does not wait for the admin message, and it allows the medicine or water to flow to the farm automatically.

Discuss briefly if the proposed system would be the suitable choice for a small to medium size Farming enterprise? What is the area to be covered by the proposed system? What would be the cost of implanting such a system? Please expand on this part.

Beaglebone Black	100$
Submersible Water Pump 2no.	50$
12v Relay 2no.	10$
GSM Module	50$
Web Camera	15$
Pipe	15$
Total	240$

The estimated cost of this workaround 250USD. If we are going to increase the better camera which covers larger area then we have to go for better water pump which can give more pressure that can cover the area covered by the cameras.

The existing method for the detection of plant disease is naked eye observation by experts by which plant disease identification and detection is achieved. To do so, there is a need for a large team of experts as well as continuous plant monitoring, which will cost very high as we do with large farms. Around the same time, farmers in some countries do not have sufficient facilities or even the knowledge that they can contact experts. Because of which expert consulting costs high as well as time-consuming too. The suggested technique is proving beneficial in tracking large crop fields under these conditions. By simply seeing the symptoms on the plant leaves, automatic detection of the diseases makes it both easier and cheaper.

Visual detection of plant disease is a more laborious and at the same timeless effective process, which can only be performed in specific areas. Whereas it will take less effort, less time and become more effective if an automatic detection technique is used.

The area covered by this system depends on how many comers we are interfacing with the BBB device and how much area each camera can cover. It can be best implemented for small-scale agricultural lands.

In this work, we are providing a database which contains healthy and unhealthy leaf. When the system is going to take a new image, it will process i.e. image processing, and compares the features which are already there in the database. Predictive analytics is a category of data analytics that aims to make predictions of future results based on historical data and analytical techniques such as statistical modeling and machine learning. The science of predictive analytics with a large degree of accuracy will produce potential insights. With the assistance of advanced methods and models for predictive analysis. Initially, we tried with around 40 images in the database the accuracy will be less. Then we increase the database size to around 400 images we got better accuracy compares to the previous database. It shows that if we have a large database, we will get better accuracy.