Automated image dataset collection system for cotton crop leaves

Project Title

Automated image dataset collection system for cotton crop leaves

Project Area of Specialization Electrical/Electronic EngineeringProject Summary

Now-a-days, automation has navigated as well as handled industrial evolution and has become far-sighted in most of the domain. But the subsidy of automation to agriculture seems at very lower scale in Pakistan. We all know dataset collection is very important for detecting the symptoms of diseases i.e. when they appear on plant leaves. The reason to collect this huge amount of dataset is that the more data you provide to the machine learning model, the faster that model can learn and improve. If this dataset collection is done by using automatic technique, then less efforts are utilized because manual collection of images of cotton crop leaves in the field is very difficult, time consuming and tedious. That is why, our purpose is to develop a semi-automated robotic system that will collect the images of cotton crop leaves in the field. Also, it will record the GPS information, soil moisture and temperature. Finally, it will store the collected images and save these to the cloud.  So, by integrating automation in the field of agriculture, efficiency and productivity can be enhanced in a numerous and multiple ways. This project will bring out the certain approaches to automate agriculture which increases irrigation, preservation of farmlands and health administration. This process is done by a semi-autonomous agriculture rover which moves around the field, collecting dataset through camera fixed on it. The images are processed using algorithms in python installer and py-charm IDE.

Project Objectives

• It will automatically collect the images of cotton crop leaves in the field.
• It will record the GPS information.
• It will record the soil moisture and temperature.
• It will store the collected images and save these to the cloud.
• It will reduce the efforts of manual dataset collection.
• It will help farmers to save their time and money without compromising on their health.
• It will provide accurate results as it is machine learning model.
• It will help solve real life problems quite efficiently and quickly regarding to disease and pests.

Project Implementation Method

Guidance systems for vehicles or implements usually consist of at least the following three parts:

1) a sensor that supplies the system with the position deviation of the vehicle or implement;
2) a controller which generates a system specific correction signal;
3) an actuator that, combined with the forward movement of the system, alters the position of the vehicle or implement.

• THE PATH PLANNER will be a DIGITAL MAP which is based on coordinate system.
• The alignment of camera is done with the help of RGB color detection technique. Camera will detect the presence of leaves then it will rotate itself either up or down or left or right, towards the target.
• The Angular Motion of Camera will be controlled by Raspberry pi.
• It will also record the Real Time Location, Soil Moisture and Environmental Condition by using different types of modules such as GPS, Humidity Sensor and Temperature Sensor.
• Then, it will store the collected images and save these to the cloud by using WI-FI.
• When all process are done then vehicle moves one step.

METHODOLOGY:

Benefits of the Project

Detection of plant leaf through some automatic technique is beneficial as it reduces efforts for manual dataset collection and also reduces large work of monitoring in big farms of crops, and at very early stage itself it detects the symptoms of diseases i.e., when they appear on plant leaves. In our project, this detection of plant leaf is completely based on image processing technique. Image processing technique provides more efficient ways to detect the plant leaves in the form of  disease that is caused by fungus, bacteria or virus on plants because mere observations by eyes to detect are not accurate. Our this system will be more cost effective, time efficient and more technically feasible so will be easily available in Pakistan. This system can be used and handled by anyone which eliminates the need of a technical instructor. This system will help farmers to save their time and money without compromising on their health. This project will contribute to the agriculture side so, by integrating automation in the field of agriculture, efficiency and productivity can be enhanced in a numerous and multiple ways.

Technical Details of Final Deliverable

The final deliverable model is to Interface the camera with raspberry pi to detect the plant that is present in the field. It will capture the images in form of main frame and it will also record the Real Time Location, Soil Moisture and Environmental Condition for using different types of modules such as GPS, Humidity Sensor, and Temperature Sensor. Then, it will store the collected images and save these to the cloud. When whole process is done then vehicle moves one step. This semi-autonomous vehicle is wirelessly connected to; ESP32 Wi-Fi, Dongle device for uploading captured images to cloud.

Final Deliverable of the Project HW/SW integrated systemCore Industry AgricultureOther IndustriesCore Technology RoboticsOther TechnologiesSustainable Development Goals 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)	67600
5MP Raspberry Pi camera module	Equipment	1	800	800
Raspberry Pi 4 model B	Equipment	1	25500	25500
Pan-tilt structure	Equipment	1	600	600
Servo motors	Equipment	3	250	750
Battery	Equipment	1	5000	5000
GPS module	Equipment	1	5500	5500
ESP32 Wi-Fi Module	Equipment	1	1600	1600
Temperature sensor	Equipment	2	100	200
Humidity sensor	Equipment	2	100	200
Micro SD card 32GB	Equipment	1	950	950
Micro HDMI to VGA Converter wire	Equipment	1	800	800
Power supply for Raspberry Pi	Equipment	1	500	500
4G Modem	Equipment	1	4000	4000
Ultrasonic sensor	Equipment	1	200	200
Vehicle Designing	Equipment	1	15000	15000
Travelling for field visit	Miscellaneous	2	3000	6000