Locust Swarm Detection and prevention of crops from locust attack using Internet of things and deep learning

Locusts are agricultural pests found in many parts of the worldwide. Globally, approximately 50 million km2 of the total land area is infested by locusts each year, and about one-eighth (1/8) of the world?s population is affected by locust plagues. In the last 2500 years, China has suffered from fre

Project Title

Locust Swarm Detection and prevention of crops from locust attack using Internet of things and deep learning

Project Area of Specialization

Computer Science

Project Summary

Locusts are agricultural pests found in many parts of the worldwide. Globally, approximately 50 million km2 of the total land area is infested by locusts each year, and about one-eighth (1/8) of the world’s population is affected by locust plagues. In the last 2500 years, China has suffered from frequent locust plagues; the extent and severity of which have been the highest in the world and attacks of locusts caused suffering of farmers in many areas of the world, especially, in Africa and sub-continent. The alarming growth and attacks of locusts are leading towards the shortage of food soon if not controlled. Researchers and scientists are tackling the situation with their ideas. Some of these ideas involve working on the possibilities of detection of locust attack in advance. However different efforts already mentioned (Use of Satellite, sensors, or chemicals) have been made to detect and prevent locusts. In addition, different stakeholders are working under NLCC to fight against locusts, but their performance is limited due to lack of effective and modern spray machines, aircraft, and drones. Meanwhile, the use of drones is still in the experimental phase but is costly. Besides these, the manual operation (spraying on locusts) is time consuming and less effective. Besides some are also working on the possibilities of detection of locust attack in advance. However, this area is left almost unexplored due to the unavailability of the dataset. Also, the set of features that may be used for the early attack detection is still ambiguous. Others are working on the deep learning-based detection of the locust swarms followed by the tackling of their attack. This project proposes an idea of tackling the attack of locusts with the help of deep learning and IoT. A camera will be deployed at a location near the field. The locust swarm is detected by the cameraa deployed in the field.  The spray bottles deployed in the field will start spraying the spray to kill the locusts to prevent the attack. In case if spray bottles have small quantity of spraying liquid available, a buzzer like sound will be produced to inform the farmers to refill the tank. Drum like music will be played once the spray tank gets empty.  It results in disturbance of locusts reasoning them to avoid the attack without killing them.

Project Objectives

Our objective is the solution to all these issues discussed above which is Apply classification on images captured by locust Spray on crops without using expensive drones, aircraft. Distract the locust using drum like music.

Notify the concerned farmers when the locust is detected. So that they can take preventive measures against locust attack and crop protection

Project Implementation Method

The project development approach for our project is described in the following phase. This is the phase division of methodology, and each section describes that how the working on the project “Locust swarm detection and prevention of crops by using deep learning and Internet of things” is done.

Benefits of the Project

The aim of this project is to design a system which ease to the farmers to prevent their crops by locust attacks. This system divides into two parts first that to detect locust which is perform by deep learning algorithm by using image dataset, we can infer that system must be detect locust swarms at a particular distance.

On the other hand, the second part of this project design IOT implementation in which we can connect devices that can perform automatic spraying on crops in which there is tank which filled with ULV formula liquid, and camera which capture locust spray and then system calculate amount of spray, we can most probably want to doing that spray can be variable amount because unlimited spraying ULV formula can damage the crops, then there is also a speaker that play drum like music in case tank of spray did not fill that time locust detect.

Technical Details of Final Deliverable

A hardware based system as a prototype named as “Locust swarm detection and prevention using deep learning and internet of things” is a deep learning and internet of things-based end to end system that is developed to provide a solution. This application is developed for detection of locust and protection of crops against locust attack. For detection of locust attack, deep learning models are used to train the model to only detect the locust and perform the protective measures against locust to protect the crops from damage. Yolo v5 model is used to train the model. Two classes namely locust and other flying insects are used to train the model.  Internet of things is used to develop the end-to-end system to automatically spray on the crops and notify the farmers whether the locust is detected, or spray tank gets empty. Different iot devices are used in the system. Such as camera will be connected to raspberry pi and servo motor. The servo motor is used to give feedback using angle by rotating the camera in 360 degrees. SIM808 GSM module sends SMS notification to owner when it detects water level lowered in water tank and when the locust is detected. Relay power is connected to raspberry pi. The relay power ups the water pump when it gets active high signal and transmit as much power to the system as needed. The active buzzer is connected to raspberry pi to produce drum like music when locust is detected, or the water tank gets empty. The spray motor connected with relay to control the power consumption. The spray motor connected to raspberry pi.

Final Deliverable of the Project

Hardware System

Core Industry

Agriculture

Other Industries

IT

Core Technology

Artificial Intelligence(AI)

Other Technologies

Internet of Things (IoT)

Sustainable Development Goals

No Poverty, Zero Hunger

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com